search![]()
Salesforce Inc. Salesforce 公司CRM 客户关系管理
Search within 搜索范围
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
close![]()
Salesforce Inc. Salesforce 公司
Interviews 采访
Panel: CRM, MSFT, NOW: AI Agents in Focus - Emerging Opportunities and Competitive Shifts
小组讨论：CRM、MSFT、NOW：聚焦 AI 代理 - 新兴机会和竞争转变
Posted January 1, 2025 发布 January 1， 2025
31 min read 31 分钟阅读
31 min read 31 分钟阅读
BizSprints 首席执行官
auto_awesome
Topics 主题AI Solutions AI 解决方案
Call Center Technology 呼叫中心技术
Customer Service 顾客服务
Data Management 数据管理
Automation 自动化
Cost Savings 节省成本
Scaling AI 扩展 AI
Implementation 实现
Summary 总结
The conversation between the Tegus Client and the expert focused on the implementation and impact of Agentforce in the Service Cloud area, with a goal of increasing efficiency, reducing remedial tasks, and improving employee satisfaction. The focus is on customer service due to its cost-saving potential, with successful automation tools in the business development rep space highlighted. The conversation also touched on the challenges companies face in modernizing their data, the timeline for proof of concepts, and the process of implementing AI agents in customer service to achieve labor savings.
Tegus 客户与专家之间的对话侧重于 Agentforce 在 Service Cloud 领域的实施和影响，目标是提高效率、减少补救任务和提高员工满意度。由于其节省成本的潜力，重点是客户服务，并强调了业务开发代表领域成功的自动化工具。对话还谈到了公司在实现数据现代化方面面临的挑战、概念验证的时间表，以及在客户服务中实施 AI 代理以实现节省劳动力的过程。
Tegus 客户与专家之间的对话侧重于 Agentforce 在 Service Cloud 领域的实施和影响，目标是提高效率、减少补救任务和提高员工满意度。由于其节省成本的潜力，重点是客户服务，并强调了业务开发代表领域成功的自动化工具。对话还谈到了公司在实现数据现代化方面面临的挑战、概念验证的时间表，以及在客户服务中实施 AI 代理以实现节省劳动力的过程。
Expert Details 专家详细信息 Compliance 合规
Chief Executive Officer at BizSprints, an MSP & VAR of different software solutions including Smartsheet. The expert is responsible for sales, integrations, customizing, and the ongoing management of those systems for their clients.
BizSprints的首席执行官，包括Smartsheet在内的各种软件解决方案的MSP和VAR。专家负责为客户销售、集成、定制和持续管理这些系统。
Interview 面试Call on December 18, 2024
2024 年 12 月 18 日致电
2024 年 12 月 18 日致电
Tegus Client Tegus 客户端
play_arrow00:00:00
Thanks for taking the time. I'm doing research on
, trying to understand better. I'm trying to understand Agentforce better as that launches and becomes reality here over the next few years. That's where I'm coming from. Maybe just give me a couple seconds of your background and the organization you lead and what your focus is.
感谢您抽出时间。我正在对
进行研究，试图更好地理解。我正在努力更好地了解 Agentforce，因为它在未来几年内在这里推出并成为现实。这就是我的家乡。也许可以给我几秒钟的背景、你领导的组织以及你的工作重点。
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:00:20
Sure. I own a few different organizations, but my background is all inside of the call center and technology space. I'm a recovering developer. I definitely know the systems both from an operational perspective, but also from a technological, what goes into these things, perspective as well. I found out actually just about two years ago, I actually was one of the first 100 customers of
. I've been around these guys forever and been working with them as both customer and partner for a long time.
确定。我拥有几个不同的组织，但我的背景都在呼叫中心和技术领域。我是一名正在康复的开发人员。我肯定从运营的角度和从技术的角度了解这些系统。实际上，就在大约两年前，我才发现，我实际上是 Salesforce 的前 100 个客户之一
。我一直和这些人在一起，作为客户和合作伙伴与他们合作了很长时间。
On the Agentforce side specifically, just to hone in on that for you, just so you know, we launched initial POCs on Agentforce. We have a total of 18 active instances right now that we're managing and collecting data off of. Like I said, I'm pretty diverse in my views of the system, both as a customer and partner and implementer. I can talk from multiple angles of it.
特别是在 Agentforce 方面，为了让您了解这一点，我们在 Agentforce 上推出了初始 POC。我们现在总共有 18 个活动实例，我们正在管理和收集数据。就像我说的，我对系统的看法非常多样化，无论是作为客户还是合作伙伴和实施者。我可以从多个角度来谈论它。
特别是在 Agentforce 方面，为了让您了解这一点，我们在 Agentforce 上推出了初始 POC。我们现在总共有 18 个活动实例，我们正在管理和收集数据。就像我说的，我对系统的看法非常多样化，无论是作为客户还是合作伙伴和实施者。我可以从多个角度来谈论它。
Tegus Client Tegus 客户端
play_arrow00:01:14
Okay. Just so I understand, you're a partner to a lot of
customers and you help them with implementations. Is that accurate?
好。据我所知，您是许多
客户的合作伙伴，您帮助他们进行实施。这准确吗？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:01:23
That's right. A couple of my other businesses that I own are actually customers of
as well. We're both customer and partner. I know it from both sides.
没错。我拥有的其他几家企业实际上
也是 的客户。我们既是客户，也是合作伙伴。我从两方面都知道。
Tegus Client Tegus 客户端
play_arrow00:01:32
Got it. Let's dig into those 18 Agentforce programs you have going on that you're seeing and working with. Maybe just keep describing those to me. What stage are all of those? What are the types of customers? What are the types of use cases? Is there a way to break it down in any of those dimensions?
明白了。让我们深入研究一下您正在查看和使用的 18 个 Agentforce 计划。也许只是继续向我描述这些。这些都是哪个阶段的？客户有哪些类型？有哪些类型的用例？有没有办法在这些维度中分解它？
明白了。让我们深入研究一下您正在查看和使用的 18 个 Agentforce 计划。也许只是继续向我描述这些。这些都是哪个阶段的？客户有哪些类型？有哪些类型的用例？有没有办法在这些维度中分解它？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:01:50
Yes. All the customers on the initial scope, and I should still say pretty much all the customers except for maybe two,
really limited it. It was all on the support side. We have some help desk clients in there. We have a couple retailers in there who are actually utilizing this for their help desk solutioning. We have a few who are doing actually their customer service aspects of it and took off a piece of their customer service teams on there and having some of the customer service solutions done.
是的。初始范围上的所有客户，我仍然应该说几乎所有客户，除了两个
， 真的限制了它。一切都在支持方面。我们有一些帮助台客户。我们那里有几家零售商，他们实际上正在将其用于他们的帮助台解决方案。我们有一些人实际上正在做他们的客户服务方面，并在那里撤走了一部分客户服务团队，并完成了一些客户服务解决方案。
Just so you understand, because of our background in BPO and business process outsourcing call center, most of what we do is, we're B2B, of course, but our customers are mostly B2C. It's mostly consumer support, except in the help desk areas. That is still B2B because the help desk, of course, is for their retail locations or remote locations or their actual staff themselves.
正如您所理解的，由于我们在 BPO 和业务流程外包呼叫中心方面的背景，我们所做的大部分工作当然是 B2B，但我们的客户主要是 B2C。它主要是消费者支持，但帮助台区域除外。这仍然是 B2B，因为服务台当然是针对他们的零售地点或偏远地点或他们的实际员工本身的。
正如您所理解的，由于我们在 BPO 和业务流程外包呼叫中心方面的背景，我们所做的大部分工作当然是 B2B，但我们的客户主要是 B2C。它主要是消费者支持，但帮助台区域除外。这仍然是 B2B，因为服务台当然是针对他们的零售地点或偏远地点或他们的实际员工本身的。
Tegus Client Tegus 客户端
play_arrow00:02:45
Got it. You actually run a call center outsourcing business?
明白了。您实际上经营着呼叫中心外包业务？
明白了。您实际上经营着呼叫中心外包业务？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:02:48
We do. We still own one. It's down in Mexico. We have an MSP and VAR as well.
我们愿意。我们仍然拥有一台。它在墨西哥。我们也有 MSP 和 VAR。
我们愿意。我们仍然拥有一台。它在墨西哥。我们也有 MSP 和 VAR。
Tegus Client Tegus 客户端
play_arrow00:02:53
Okay. In these cases, these pilots you're talking about, are they for customers who are running their own customer service centers or these are pilots for customers who are going to be your BPO customers and you're implementing Agentforce in your environment for them?
好。在这些情况下，您谈论的这些试点是针对运行自己的客户服务中心的客户，还是针对将成为您的 BPO 客户的客户的试点，并且您正在您的环境中为他们实施 Agentforce？
好。在这些情况下，您谈论的这些试点是针对运行自己的客户服务中心的客户，还是针对将成为您的 BPO 客户的客户的试点，并且您正在您的环境中为他们实施 Agentforce？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:03:09
There's a mixture in that answer, to be honest with you. Just to understand our world, all of our customers that I'm talking about, those instances are managed by our MSP and VAR. Some of those customers also happen to be customers of my BPO as well. There's a little bit of a combo platter there. I will tell you, the mass majority, if I look across the 18 instances we have, all but eight of them, so 10 of the 18 are customers of our BPO as well.
老实说，这个答案是混合的。为了了解我们的世界，我所说的所有客户，这些实例都由我们的 MSP 和 VAR 管理。其中一些客户恰好也是我的 BPO 的客户。那里有一点组合拼盘。如果我查看我们拥有的 18 个实例，除了 8 个实例之外，其他所有实例都是大多数，因此 18 个实例中有 10 个也是我们 BPO 的客户。
老实说，这个答案是混合的。为了了解我们的世界，我所说的所有客户，这些实例都由我们的 MSP 和 VAR 管理。其中一些客户恰好也是我的 BPO 的客户。那里有一点组合拼盘。如果我查看我们拥有的 18 个实例，除了 8 个实例之外，其他所有实例都是大多数，因此 18 个实例中有 10 个也是我们 BPO 的客户。
The majority of what they're looking to do, just so you understand, and this probably gives you the keys to the kingdom here, what most businesses are trying to do is maximize utilization and increase utilization of their internal staff and also to remove what can be to be able to also potentially outsource it more. What I mean by that is, let's say they have 200 staff internal still, what they'd like to do is most internal call center staff or support staff, whatever it is. I'm saying call center just because of my experience with that, but it could be anything. It could be IT, it could be HR, it could be whatever.
他们想要做的大部分事情，只是为了让你明白，这可能给你通往这里王国的钥匙，大多数企业试图做的是最大限度地提高利用率和提高内部员工的利用率，并消除可能的东西，以便也有可能将其外包更多。我的意思是，假设他们内部仍有 200 名员工，他们想做的是大多数内部呼叫中心员工或支持人员，无论是什么。我说呼叫中心只是因为我在这方面的经验，但它可以是任何东西。可能是 IT，可能是 HR，可以是任何东西。
他们想要做的大部分事情，只是为了让你明白，这可能给你通往这里王国的钥匙，大多数企业试图做的是最大限度地提高利用率和提高内部员工的利用率，并消除可能的东西，以便也有可能将其外包更多。我的意思是，假设他们内部仍有 200 名员工，他们想做的是大多数内部呼叫中心员工或支持人员，无论是什么。我说呼叫中心只是因为我在这方面的经验，但它可以是任何东西。可能是 IT，可能是 HR，可以是任何东西。
Internal staff usually runs around 60% utilization. If I have an hour of a person's time, I'm only getting about 31-35 minutes out of an average hour of work production. Even that productional work is only done at an efficiency rate of around 40%. What that means is an average employee who works directly for a business will only produce actual quantifiable work that is productive around about 20 minutes per hour per day. In contrast to an outsourcer, so an HRO, ITO, BPO, whatever, we produce on average around 40-42 minutes of production-ready work per hour. We're double.
内部员工的利用率通常约为 60%。如果我有一个小时的时间，那么我平均一小时的工作制作中只能得到大约 31-35 分钟。即使是这些生产工作也只能以 40% 左右的效率完成。这意味着，直接为企业工作的普通员工每天每小时只能生产大约 20 分钟的可量化工作。与外包商（HRO、ITO、BPO 等）相比，我们平均每小时生产大约 40-42 分钟的生产就绪工作。我们是双重的。
内部员工的利用率通常约为 60%。如果我有一个小时的时间，那么我平均一小时的工作制作中只能得到大约 31-35 分钟。即使是这些生产工作也只能以 40% 左右的效率完成。这意味着，直接为企业工作的普通员工每天每小时只能生产大约 20 分钟的可量化工作。与外包商（HRO、ITO、BPO 等）相比，我们平均每小时生产大约 40-42 分钟的生产就绪工作。我们是双重的。
What companies are typically looking to do is their goal with AI and automationing is to be able to increase efficiency and productional workload from around that 20-22 minutes per hour up to around 30-35 minutes per hour and/or reduce the busy work that they still have internally to remove that completely gone to actually automate it or move more of that back out into the outsourcer by being able to utilize these AI tools to allow the outsourcer actually braille perform those tasks.
公司通常希望做的是，他们通过 AI 和自动化的目标是能够将效率和生产工作量从每小时 20-22 分钟左右提高到每小时 30-35 分钟左右，和/或减少他们内部仍然需要的繁忙工作，以完全消除这些工作，以实际自动化它，或者通过能够利用这些 AI 将更多工作移回外包商允许外包商实际盲文执行这些任务的工具。
公司通常希望做的是，他们通过 AI 和自动化的目标是能够将效率和生产工作量从每小时 20-22 分钟左右提高到每小时 30-35 分钟左右，和/或减少他们内部仍然需要的繁忙工作，以完全消除这些工作，以实际自动化它，或者通过能够利用这些 AI 将更多工作移回外包商允许外包商实际盲文执行这些任务的工具。
Tegus Client Tegus 客户端
play_arrow00:05:22
Yes. The vision is that these artificial intelligence tools will help take some of the simpler use cases and allow that that are viewed as lower-value time for those employees to be focused on those and they're maybe inefficient in doing it, move those over to the more efficient either automate them or more efficient BPO and allow and just try to leave the highest value, most complex use cases with the employees. Is that fair?
是的。愿景是，这些人工智能工具将有助于处理一些更简单的用例，并允许那些被视为低价值的时间让这些员工专注于这些，而他们可能在这样做时效率低下，将这些转移到更高效的自动化或更高效的 BPO 上，并允许并尝试留下最高价值， 员工最复杂的用例。这公平吗？
是的。愿景是，这些人工智能工具将有助于处理一些更简单的用例，并允许那些被视为低价值的时间让这些员工专注于这些，而他们可能在这样做时效率低下，将这些转移到更高效的自动化或更高效的 BPO 上，并允许并尝试留下最高价值， 员工最复杂的用例。这公平吗？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:05:49
That's exactly right. There's multiple reasons for that too. It's not even just from a dollars and cents efficiency perspective, which is the most apparent from a financial modeling, but also you have to take the secondary piece of this which is decades of research which will show that if you want to truly affect your EPS score, your employee satisfaction scores internally for satisfaction of work. The way to do that is reduction of remedial or repetitive tasking and increasing of high complex, what you'd consider higher stress tasking.
这是完全正确的。这也有多种原因。这甚至不仅仅是从金钱和美分的效率角度来看，这在财务建模中最为明显，而且你还必须采取第二部分，即数十年的研究，这将表明，如果你想真正影响你的 EPS 分数，你的员工满意度在内部对工作满意度的评分。做到这一点的方法是减少补救性或重复性任务，增加高复杂性，即你认为的更高压力任务。
这是完全正确的。这也有多种原因。这甚至不仅仅是从金钱和美分的效率角度来看，这在财务建模中最为明显，而且你还必须采取第二部分，即数十年的研究，这将表明，如果你想真正影响你的 EPS 分数，你的员工满意度在内部对工作满意度的评分。做到这一点的方法是减少补救性或重复性任务，增加高复杂性，即你认为的更高压力任务。
As humans, we're naturally tuned to want to handle more complex things because the outcomes of those is typically more satisfying to us from a work perspective. The more we can reduce those remedial tasks, we're actually improving employee satisfaction rates, which of course, will then decrease churn rates, decrease requirements for additional training and new hire training, etc.
作为人类，我们自然而然地想要处理更复杂的事情，因为从工作的角度来看，这些结果通常更让我们满意。我们越能减少这些补救任务，我们实际上是在提高员工满意度，当然，这将降低流失率，减少对额外培训和新员工培训的要求等。
作为人类，我们自然而然地想要处理更复杂的事情，因为从工作的角度来看，这些结果通常更让我们满意。我们越能减少这些补救任务，我们实际上是在提高员工满意度，当然，这将降低流失率，减少对额外培训和新员工培训的要求等。
There's a lot more impact than just the direct one-to-one costing. That's why though
is getting nowhere near the $2 per transactional cost they want, it's fully justifiable. I could run you through a math exercise that can justify it 1,000 times over. My point is that, definitely, there is a direct correlation to companies focusing on removing initially, and that is where we're seeing all the case study and all the POCs that are running right now is remedial transactional tasking and working towards pushing that into Agentforce initially.
影响远不止直接的一对一成本核算。这就是为什么尽管
的每笔交易成本远未达到他们想要的 2 美元，但这是完全合理的。我可以带你做一个数学练习，可以证明它有 1,000 倍的合理性。我的观点是，毫无疑问，这与最初专注于删除的公司有直接关系，这就是我们看到的所有案例研究和现在正在运行的所有 POC 都是补救易任务，并努力将其最初推送到 Agentforce。
Tegus Client Tegus 客户端
play_arrow00:07:21
Okay. I'll come back to a lot of that, I'm sure, but let me just make sure I understand. Eight of the 18 pilots, but given your history, a lot of it is focused on Service Cloud and a broad definition of customer service whether that's internal help desk customer service or external B2C customer service. Outside of your organization, do you think there's a lot of pilots in the sales and marketing and commerce areas of Agentforce or this Service Cloud customer service use case is really where Agentforce's focus is today?
好。我敢肯定，我会回到很多内容，但让我确保我理解。18 个试点项目中有 8 个，但根据您的历史，其中很多都集中在 Service Cloud 和客户服务的广泛定义上，无论是内部帮助台客户服务还是外部 B2C 客户服务。在您的组织之外，您认为 Agentforce 的销售和营销以及商务领域是否有很多试点，或者这个 Service Cloud 客户服务用例确实是 Agentforce 今天的重点所在？
好。我敢肯定，我会回到很多内容，但让我确保我理解。18 个试点项目中有 8 个，但根据您的历史，其中很多都集中在 Service Cloud 和客户服务的广泛定义上，无论是内部帮助台客户服务还是外部 B2C 客户服务。在您的组织之外，您认为 Agentforce 的销售和营销以及商务领域是否有很多试点，或者这个 Service Cloud 客户服务用例确实是 Agentforce 今天的重点所在？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:07:53
That's where Agentforce is focused, in on Service Cloud. Actually, I don't even think it's plugged into Sales Cloud yet.
这就是 Agentforce 专注于 Service Cloud 的地方。实际上，我什至不认为它已插入 Sales Cloud。
这就是 Agentforce 专注于 Service Cloud 的地方。实际上，我什至不认为它已插入 Sales Cloud。
Tegus Client Tegus 客户端
play_arrow00:08:00
Is your sense that stuff will obviously come over time? From the customer base, is there a lot of interest, you think, also in those use cases or everyone wants to get the customer service use case done first?
你觉得东西显然会随着时间的推移而出现吗？从客户群来看，您认为是否对这些用例也有很多兴趣，或者每个人都想先完成客户服务用例？
你觉得东西显然会随着时间的推移而出现吗？从客户群来看，您认为是否对这些用例也有很多兴趣，或者每个人都想先完成客户服务用例？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:08:13
Yes. Just to answer, just so I can give clarity, it is available on Sales Cloud now. The Service Cloud is the holy grail. The reality is that there's so many tools out there who are already making impacts. Listen, if you can get sales all figured out and AId, great, but the reality is there's a lot of data showing just the reluctancy of the consumer market right now on doing those transactions in the bigger transactional especially.
是的。只是为了回答，为了澄清一下，它现在可以在 Sales Cloud 上使用。Service Cloud 是圣杯。现实情况是，市面上有很多工具已经产生了影响。听着，如果你能把销售全部弄清楚并且 AId，那就太好了，但现实情况是，有很多数据表明，现在消费市场不愿意进行这些交易，尤其是在更大的交易中。
是的。只是为了回答，为了澄清一下，它现在可以在 Sales Cloud 上使用。Service Cloud 是圣杯。现实情况是，市面上有很多工具已经产生了影响。听着，如果你能把销售全部弄清楚并且 AId，那就太好了，但现实情况是，有很多数据表明，现在消费市场不愿意进行这些交易，尤其是在更大的交易中。
On the Service Cloud, the reason that's so impactful is because still to this day, let's face it, support is a cost center. No matter how you want to shake it out, no matter how many times I've been on stage even telling people, "Oh no, it's a revenue center, " that's BS. Let's just get down to financial nuts and bolts. It's a cost center. Anything you can do to reduce and remove it is just adding points to your bottom line in massive, massive ways. It is the true holy grail of this whole agent movement. It is why there's such a big focus there.
在 Service Cloud 上，之所以如此有影响力，是因为直到今天，让我们面对现实吧，支持仍然是一个成本中心。无论你想怎么摇晃它，无论我在舞台上多少次甚至告诉人们，“哦，不，这是一个收入中心，”那是 BS。让我们开始讨论财务的具体细节。它是一个成本中心。您可以做的任何事情来减少和消除它，都只是以大量的方式为您的底线增加分数。它是整个代理运动的真正圣杯。这就是为什么那里有如此大的关注点。
在 Service Cloud 上，之所以如此有影响力，是因为直到今天，让我们面对现实吧，支持仍然是一个成本中心。无论你想怎么摇晃它，无论我在舞台上多少次甚至告诉人们，“哦，不，这是一个收入中心，”那是 BS。让我们开始讨论财务的具体细节。它是一个成本中心。您可以做的任何事情来减少和消除它，都只是以大量的方式为您的底线增加分数。它是整个代理运动的真正圣杯。这就是为什么那里有如此大的关注点。
The other thing, and which I did say too, is that is there use cases in other aspects? Absolutely. We're already seeing huge impacts even from external systems with that. For instance, in the BDR world, in SaaS, BDRs are business development reps who are actually doing the lead generation, we've already seen agent systeming that's been coming out for years now that has just decimated the BDR rules. We have, for instance, a case study where we are working with a team.
另一件事，我也说过，在其他方面是否有用例？绝对。我们已经看到了来自外部系统的巨大影响。例如，在 BDR 领域，在 SaaS 中，BDR 是实际负责潜在客户生成的业务开发代表，我们已经看到多年来出现的代理系统刚刚摧毁了 BDR 规则。例如，我们有一个案例研究，我们与一个团队合作。
另一件事，我也说过，在其他方面是否有用例？绝对。我们已经看到了来自外部系统的巨大影响。例如，在 BDR 领域，在 SaaS 中，BDR 是实际负责潜在客户生成的业务开发代表，我们已经看到多年来出现的代理系统刚刚摧毁了 BDR 规则。例如，我们有一个案例研究，我们与一个团队合作。
They had around 450 business development rep agents on their team. We applied some third-party systems to be able to utilize AI and some machine learning systems to be able to automate the lead development roles. That team today is only 84 agents now from over 470 and they're performing almost 3X the amount of lead creation that the larger 400+ team did. That's just one of hundreds of case studies like that.
他们的团队中大约有 450 名业务开发代表代理。我们应用了一些第三方系统来利用 AI 和一些机器学习系统来自动化首席开发角色。今天，该团队现在只有 84 名代理，来自 470 多名，他们执行的潜在客户创建量几乎是 400+ 个大型团队的 3 倍。这只是数百个类似案例研究中的一个。
他们的团队中大约有 450 名业务开发代表代理。我们应用了一些第三方系统来利用 AI 和一些机器学习系统来自动化首席开发角色。今天，该团队现在只有 84 名代理，来自 470 多名，他们执行的潜在客户创建量几乎是 400+ 个大型团队的 3 倍。这只是数百个类似案例研究中的一个。
In the business development world and especially in lead gen, it's being done. It's already there. It's proven. We know it exists. The customer service side is just so much more complex and it's so much more versatile. Every engagement is so different than the next and that is really what's pushing, I think, a lot of the companies to focus here because the ROIs on it also are massive.
在业务发展领域，尤其是在潜在客户开发领域，它正在被完成。它已经在那里了。这是经过验证的。我们知道它存在。客户服务方面要复杂得多，而且用途也要多得多。每一次参与都与下一次互动截然不同，我认为，这确实是推动许多公司关注这里的原因，因为它的投资回报率也很高。
在业务发展领域，尤其是在潜在客户开发领域，它正在被完成。它已经在那里了。这是经过验证的。我们知道它存在。客户服务方面要复杂得多，而且用途也要多得多。每一次参与都与下一次互动截然不同，我认为，这确实是推动许多公司关注这里的原因，因为它的投资回报率也很高。
Tegus Client Tegus 客户端
play_arrow00:10:34
Got it. On the tools for automation in the business development rep space, is that stuff you think most
customers get from Salesforce or there's just, you're saying, a plethora of smaller point product companies out there doing things to help?
明白了。关于业务开发代表领域的自动化工具，您认为大多数
客户是从 Salesforce 获得的东西，还是您说，有大量较小的单点产品公司在提供帮助？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:10:49
The list of companies who started there is just endless. It really is. I could sit here and spend the next hour just giving you names of those companies alone. I think
has a great opportunity to probably have a lot of those companies convert back into Agentforce, and I think they will. You'll start seeing those coming. The reality is, it's funny,
made popular the whole BDR, SDR, AE motioning in the SaaS world.
从那里开始的公司名单是无穷无尽的。确实如此。我可以坐在这里，在接下来的一个小时里只告诉你这些公司的名字。我认为
有一个很好的机会，可能会让很多这样的公司转换回 Agentforce，我认为他们会的。你会开始看到这些的到来。现实情况是，有趣的是，
使整个 BDR、SDR、AE 在 SaaS 世界中流行起来。
It only makes sense that now with Agentforce coming out, I guarantee, you will see more and more companies who are
businesses who use that same model for their sales start using Agentforce probably for their BDR efforting. As I mentioned, what we've seen more popular right now is the service side. Even from our inquiries coming in, a lot of it is from the support and service side.
我保证，现在随着 Agentforce 的推出，您会看到越来越多的
企业公司使用相同的销售模型，他们开始使用 Agentforce，这可能是为了他们的 BDR 工作。正如我所提到的，我们现在看到更流行的是服务端。即使从我们收到的询问中，也很多来自支持和服务方面。
Tegus Client Tegus 客户端
play_arrow00:11:40
The end mission and goal with all these Service Cloud Agentforce trials, is it ultimately to reduce and replace labor costs, or are there a lot of people who are doing this thinking with their end goal being better service, higher retention rates, higher NPS, things focused on that, or is it really all focused on automating out labor costs?
所有这些 Service Cloud Agentforce 试用的最终任务和目标，最终是减少和取代劳动力成本，还是有很多人在进行这种思考，他们的最终目标是更好的服务、更高的保留率、更高的 NPS，专注于此，或者真的都专注于自动化劳动力成本？
所有这些 Service Cloud Agentforce 试用的最终任务和目标，最终是减少和取代劳动力成本，还是有很多人在进行这种思考，他们的最终目标是更好的服务、更高的保留率、更高的 NPS，专注于此，或者真的都专注于自动化劳动力成本？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:12:01
I would tell you that there's a little bit of both in there. I think what is spoken out loud is much increase and improve performance and customer experience. What's driving it is labor reduction. This is where I will tell you, the one thing that
has 1,000% done right with how they have gone about their model is the one thing that I've been screaming from the rooftops, from everywhere, is that they went consumption because that is exactly how this needed to be done.
我想告诉你，两者兼而有之。我认为大声说出来的是大大提高和改善性能和客户体验。推动它的是劳动力减少。这就是我要告诉你的地方，
在他们如何实施他们的模型方面 1,000% 做得正确的一件事是，我一直在屋顶上，从四面八方尖叫的一件事是，他们去消费了，因为这正是需要做的事情。
A licensing model makes zero sense in this environment. Consumptional modeling is the only way that makes sense. When you look from just a revenue generational perspective, when we actually done the modeling on this, the great part is that as we've already started seeing some of our clients now, our early clients who we launched, the beauty is that we're already seeing the models working to where the consumptional based modelings actually will even exceed the licensing revenues. We see it about nine months to where we start seeing enough of it to where we can see positive upside beyond what we would have seen with licensing.
在这种环境中，许可模式毫无意义。消费建模是唯一有意义的方法。当你只从创收的角度来看时，当我们真正对此进行建模时，最重要的部分是，我们现在已经开始看到我们的一些客户，我们推出的早期客户，美妙的是，我们已经看到基于消费的模型实际上甚至会超过许可收入。我们看到它大约九个月，我们开始看到足够的它，以至于我们可以看到超出我们在许可方面看到的积极影响。
在这种环境中，许可模式毫无意义。消费建模是唯一有意义的方法。当你只从创收的角度来看时，当我们真正对此进行建模时，最重要的部分是，我们现在已经开始看到我们的一些客户，我们推出的早期客户，美妙的是，我们已经看到基于消费的模型实际上甚至会超过许可收入。我们看到它大约九个月，我们开始看到足够的它，以至于我们可以看到超出我们在许可方面看到的积极影响。
Tegus Client Tegus 客户端
play_arrow00:13:11
Meaning you're seeing customers could be spending more on Agentforce than they are on the Service Cloud seats?
这意味着您看到客户在 Agentforce 上的支出可能比在 Service Cloud 席位上的支出更多？
这意味着您看到客户在 Agentforce 上的支出可能比在 Service Cloud 席位上的支出更多？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:13:18
General license. You got it. That's exactly right.
通用许可证。没问题。这是完全正确的。
通用许可证。没问题。这是完全正确的。
Tegus Client Tegus 客户端
play_arrow00:13:21
How does the math usually end up working? I get that the pricing is $2 as a starting point and it's all over the place based on different types of customers. Walk me through it in this way because it's the way I'm trying to think about it. A service rep in terms of labor costs, I don't know, $30,000, $40,000, $50,000, or whatever, how much is the labor cost? How much is the Service Cloud human seat historically?
数学通常最终是如何运作的？我知道定价是 2 美元作为起点，而且根据不同类型的客户，它无处不在。以这种方式引导我完成它，因为这是我试图思考它的方式。就人工成本而言，我不知道服务代表，30,000 美元、40,000 美元、50,000 美元，或者其他任何，人工成本是多少？Service Cloud 人类席位历史上有多少钱？
数学通常最终是如何运作的？我知道定价是 2 美元作为起点，而且根据不同类型的客户，它无处不在。以这种方式引导我完成它，因为这是我试图思考它的方式。就人工成本而言，我不知道服务代表，30,000 美元、40,000 美元、50,000 美元，或者其他任何，人工成本是多少？Service Cloud 人类席位历史上有多少钱？
If you are replacing that amount of work that a single person was doing and getting rid of that whatever couple thousand dollars that
was getting for that human seat, what is that replaced with? Is it replaced with twice, 3X, 4X, 5X as much Agentforce dollars? Maybe walk me through how you think about that math really from a
perspective or from a customer's perspective.
如果你要替换一个人所做的工作量，并摆脱
为这个人工席位获得的几千美元，那么它被什么取代呢？它是否被两倍、3 倍、4 倍、5 倍的 Agentforce 美元所取代？也许可以跟我说说，你是如何从
或客户的角度来思考这个数学的。
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:14:11
Sure. Let me just give you a quick breakdown of some numbers here. These are the keys to the kingdom here. It's easiest to do with BPO because our rates are pretty averageable and they're pretty understood and it sets the tone for especially the industry. The offshoring labor rate averages about $11.25 per hour per agent. LATAM is averaging around about, it's a little bit less, but we'll just call it about $14 per hour per agent. Domestic, this varies quite a bit, but it's usually around $36 per hour. The reason that's important is because how you then go from there.
确定。让我在这里给你简单介绍一下一些数字。这些是通往这里王国的钥匙。使用 BPO 是最容易做到的，因为我们的费率相当平均，而且他们很容易理解，这尤其为行业定下了基调。离岸劳动力费率平均为每个代理每小时约 11.25 美元。LATAM 的平均价格大约低一点，但我们只称每个代理每小时 14 美元左右。国内，差异很大，但通常在每小时 36 美元左右。这很重要的原因是你如何从那里开始。
确定。让我在这里给你简单介绍一下一些数字。这些是通往这里王国的钥匙。使用 BPO 是最容易做到的，因为我们的费率相当平均，而且他们很容易理解，这尤其为行业定下了基调。离岸劳动力费率平均为每个代理每小时约 11.25 美元。LATAM 的平均价格大约低一点，但我们只称每个代理每小时 14 美元左右。国内，差异很大，但通常在每小时 36 美元左右。这很重要的原因是你如何从那里开始。
First thing you need to understand is that the benchmark for all sourcers globally is that we work towards being at what we would consider about 70% utilizational rates. What that means is if you figure that at 60 minutes, of course, in an hour, and if I could provide 70% utilization or build what we considered active productional time, that's about 42 minutes an hour, which is decent. Just for this math exercise, let's use the lowest. We'll use the offshore rates. At $11.25, if you divide that out over 42 minutes, you come out to around $0.27 per minute of cost.
首先，您需要了解的是，全球所有采购商的基准是，我们努力达到我们认为的 70% 左右的利用率。这意味着，如果你计算一下 60 分钟，当然是一小时，如果我能提供 70% 的利用率或建立我们认为的活跃生产时间，那大约是每小时 42 分钟，这是相当不错的。对于这个数学练习，让我们使用最低的。我们将使用离岸费率。如果将其除以 42 分钟，则每分钟的成本约为 0.27 USD。
首先，您需要了解的是，全球所有采购商的基准是，我们努力达到我们认为的 70% 左右的利用率。这意味着，如果你计算一下 60 分钟，当然是一小时，如果我能提供 70% 的利用率或建立我们认为的活跃生产时间，那大约是每小时 42 分钟，这是相当不错的。对于这个数学练习，让我们使用最低的。我们将使用离岸费率。如果将其除以 42 分钟，则每分钟的成本约为 0.27 USD。
Now, with that then figured, if you figured that the average tier one customer service interaction takes around 5.5 minutes, and that could be low or high, depending on the type of company. Certain industries are around 15. Others can be up to a half hour. Let's just use the 5.5. At 5.5 minutes, you're at now about $1.48 or $1.49 per interaction at rough cost. That doesn't even include all my overheads and soft costs that go in on top of that.
现在，考虑到这一点，如果您认为平均一级客户服务交互大约需要 5.5 分钟，这可能是低或高，具体取决于公司的类型。某些行业大约是 15 岁。其他可能长达半小时。让我们只使用 5.5。在 5.5 分钟时，您现在每次互动的费用约为 1.48 USD 或 1.49 USD（大致成本）。这甚至不包括我所有的管理费用和软成本。
现在，考虑到这一点，如果您认为平均一级客户服务交互大约需要 5.5 分钟，这可能是低或高，具体取决于公司的类型。某些行业大约是 15 岁。其他可能长达半小时。让我们只使用 5.5。在 5.5 分钟时，您现在每次互动的费用约为 1.48 USD 或 1.49 USD（大致成本）。这甚至不包括我所有的管理费用和软成本。
At $1.485, I would add in here at 1.33%, which brings it to $1.97. There's also variable cost on there, so you'd actually sometimes even add a little bit more. That rough math, just so you understand, is how
got to its $2 per interaction. It's exactly that what I just walked you through. The moment I saw $2, I already knew that number because that's the number we use. That is just outsourcing 101 for offshoring.
在 1.485 美元时，我会以 1.33% 的价格加入这里，从而达到 1.97 美元。那里也有可变成本，所以你有时甚至会多加一点。为了让您理解，这个粗略的计算就是
如何达到每次互动 2 美元的原因。这正是我刚刚带你经历的。当我看到 2 美元的那一刻，我已经知道这个数字了，因为这是我们使用的数字。那只是将 101 外包给离岸外包。
If you do that same exact math exercise I just walked you through domestically, as you can imagine, it's basically 3X higher. It's almost $6 per interaction that you're paying, if not higher, to be honest with you. That, mind you, is with outsourcing numbers. If you go in-house and you use captive call centers or call centers that are run by the business, these numbers get exponentially crazier.
如果你做我刚才在国内带你完成的完全相同的数学练习，你可以想象，它基本上高出 3 倍。老实说，您每次支付的互动几乎是 6 美元，如果不是更高的话。请注意，这是与外包数字有关的。如果您进入内部并使用专属呼叫中心或由企业运营的呼叫中心，这些数字会呈指数级增长。
如果你做我刚才在国内带你完成的完全相同的数学练习，你可以想象，它基本上高出 3 倍。老实说，您每次支付的互动几乎是 6 美元，如果不是更高的话。请注意，这是与外包数字有关的。如果您进入内部并使用专属呼叫中心或由企业运营的呼叫中心，这些数字会呈指数级增长。
Let's just give you an example here, near shore. We'll just use near shore. Even a near shore internal captive call centers are still paying around $16 per hour per agent. I'm giving them a lot, to be honest, because it's actually over $20, but let's just say it's $16. Where things get crazy is utilizational rate. Utilization rates of a captive agent are normally around 56%.
我们在这里举个例子，靠近海岸。我们只使用近岸。即使是近岸内部专属呼叫中心，每个座席每小时也要支付 16 美元左右。老实说，我给他们很多，因为实际上超过 20 美元，但让我们只说它是 16 美元。事情变得疯狂的地方是利用率。专属代理的利用率通常在 56% 左右。
我们在这里举个例子，靠近海岸。我们只使用近岸。即使是近岸内部专属呼叫中心，每个座席每小时也要支付 16 美元左右。老实说，我给他们很多，因为实际上超过 20 美元，但让我们只说它是 16 美元。事情变得疯狂的地方是利用率。专属代理的利用率通常在 56% 左右。
Though, if you take that and you figure at 56%, you're only getting production time of around 34 minutes per hour. At $16, and you divide that out by 34, I'm now at $0.47 per minute. If I times that out now by my 5.5, I'm already at $2.60 per call, and it gets a lot crazier, a lot faster. Captive centers are the biggest target for this because I could charge $4, to be honest with you, per interaction and still show savings to companies if I had the opportunity.
不过，如果你把这个数字计算在 56% 左右，你只能得到每小时大约 34 分钟的制作时间。在 16 USD 中，除以 34，我现在每分钟 0.47 USD。如果我现在用 5.5 加倍，我每次通话已经是 2.60 美元了，而且它会变得更疯狂、更快。自保中心是最大的目标，因为老实说，我可以在每次互动中收取 4 美元的费用，并且如果有机会，我仍然可以向公司展示节省的费用。
不过，如果你把这个数字计算在 56% 左右，你只能得到每小时大约 34 分钟的制作时间。在 16 USD 中，除以 34，我现在每分钟 0.47 USD。如果我现在用 5.5 加倍，我每次通话已经是 2.60 美元了，而且它会变得更疯狂、更快。自保中心是最大的目标，因为老实说，我可以在每次互动中收取 4 美元的费用，并且如果有机会，我仍然可以向公司展示节省的费用。
Tegus Client Tegus 客户端
play_arrow00:17:43
When Salesforce goes through that math with customers, how much of that value is Salesforce capturing?
当 Salesforce 与客户进行数学计算时，Salesforce 捕获了多少价值？
当 Salesforce 与客户进行数学计算时，Salesforce 捕获了多少价值？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:17:49
Here's the problem. Number one is
doesn't understand the math right now, really. That's why they're having to hire in a whole new breed of AEs because their account executives who are in the seats right now have no conception of consumption at all. I've talked to them. These guys have zero clue. Unfortunately, a lot of their leadership doesn't even understand it, to be honest with you. The people who understand it is the partners like me and people who come from not that world.
这就是问题所在。首先是
现在真的不懂数学。这就是为什么他们不得不招聘一种全新的 AE，因为他们现在坐在座位上的客户经理根本没有消费的概念。我已经和他们谈过了。这些家伙毫无头绪。不幸的是，老实说，他们的很多领导甚至不理解这一点。理解它的人是像我这样的合作伙伴和来自那个世界以外的人。
Just to give you example, my average customer is only paying $0.50 or $0.60 per transaction right now, as they should be though too. I'm not saying everything's sunshine and gloriousness with Agentforce. There's some issues in there too. Most of the reason is because
themselves does not know how to sell this model. They need to go out and find the old telecom guys and bring them in, but they haven't done that yet. They can't even accurately articulate the ROI.
举个例子，我的普通客户现在每笔交易只支付 0.50 美元或 0.60 美元，他们也应该这样做。我并不是说 Agentforce 的一切都是阳光和光荣的。这里面也有一些问题。大部分原因是
自己不知道如何销售这种模式。他们需要出去找到老的电信公司，把他们带进来，但他们还没有这样做。他们甚至无法准确地阐明 ROI。
Now, I can tell you how we do it because we're of course having these conversations. For us, it's easy to articulate that math and walk a person back through to it. The questions we're getting asked and pushback, which is the right questions and we'll even call them out is, for instance, bounce rates. Bounce rates right now are still a little bit higher than I know
would like them to be and that we would like them to be.
现在，我可以告诉你我们是怎么做的，因为我们当然正在进行这些对话。对我们来说，很容易阐明这个数学并引导人们回顾它。我们被问到的问题和反对意见，这是正确的问题，我们甚至会指出来，例如，跳出率。现在的跳出率仍然比我知道
希望的和我们希望的
要高一点。Bounce rates are the percentage of volume that is leaving the agent and then going back out still to a live agent for additional support because the actual virtual agent wasn't able to fully support them or hit a limiter. Right now, we're still seeing averaging, I'm rounding it, but it's about 30-ish percent. 25%-30% of volume that's going into it is bouncing out to an actual live agent still.
退回邮件率是指由于实际的虚拟代理无法完全支持他们或达到限制而离开代理，然后仍然返回给实时代理以获得额外支持的数量的百分比。现在，我们仍然看到平均值，我四舍五入，但大约是 30%。进入它的 25%-30% 的流量仍然会反弹到真正的现场代理。
退回邮件率是指由于实际的虚拟代理无法完全支持他们或达到限制而离开代理，然后仍然返回给实时代理以获得额外支持的数量的百分比。现在，我们仍然看到平均值，我四舍五入，但大约是 30%。进入它的 25%-30% 的流量仍然会反弹到真正的现场代理。
Tegus Client Tegus 客户端
play_arrow00:19:39
Okay. That is viewed as not that good? That sounds pretty good. That means they're able to deflect 60% or 70% of them, no?
好。那被认为是不是那么好吗？这听起来不错。这意味着他们能够偏转其中的 60% 或 70%，不是吗？
好。那被认为是不是那么好吗？这听起来不错。这意味着他们能够偏转其中的 60% 或 70%，不是吗？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:19:47
Yes, it is. It's both sides. The short answer is, yes, that's great. The bad answer is it's still not that good. Mind you, this is defined traffic. This isn't just general traffic. This isn't like I send everything and I only get 20% or 30% back. All of these POCs are defined traffic, which means I've specified a specific engagement types. It could be social, it could be text, it could be chat, could be whatever. I defined a certain segment that I'm pushing to this, and of that segment, I'm supporting 60%, 70%. Where in an IVR or even some certain, web chat systems and everything, my bounce-out rates are only 15%-20%. It's still higher than the industry average.
是的，它是。这是两边的。简短的回答是，是的，这很好。糟糕的回答是它仍然没有那么好。请注意，这是定义的流量。这不仅仅是一般流量。这并不是说我发送了所有东西，但我只收到了 20% 或 30% 的退款。所有这些 POC 都是定义的流量，这意味着我已经指定了特定的参与类型。它可以是社交的，可以是文本的，可以是聊天的，可以是任何东西。我定义了一个特定的细分市场，我将其推送到这个细分市场，在该细分市场中，我支持 60%、70%。在 IVR 甚至某些网络聊天系统和所有系统中，我的跳出率只有 15%-20%。它仍然高于行业平均水平。
是的，它是。这是两边的。简短的回答是，是的，这很好。糟糕的回答是它仍然没有那么好。请注意，这是定义的流量。这不仅仅是一般流量。这并不是说我发送了所有东西，但我只收到了 20% 或 30% 的退款。所有这些 POC 都是定义的流量，这意味着我已经指定了特定的参与类型。它可以是社交的，可以是文本的，可以是聊天的，可以是任何东西。我定义了一个特定的细分市场，我将其推送到这个细分市场，在该细分市场中，我支持 60%、70%。在 IVR 甚至某些网络聊天系统和所有系统中，我的跳出率只有 15%-20%。它仍然高于行业平均水平。
Tegus Client Tegus 客户端
play_arrow00:20:29
Can we go back to the math? That was super helpful to understand the cost per human conversation. What do you end up spending per seat with
for customers for Service Cloud in the pre-Agentforce world?
我们能回到数学上来吗？这对于了解每次人类对话的成本非常有帮助。
在 Agentforce 出现之前
，您最终为 Service Cloud 客户提供 的每个席位花费多少？Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:20:42
In the pre Agentforce world, on the Service Cloud side, it's averaging around $275 per seat. $225-$275 once you get all the stuff loaded in there.
在 Agentforce 之前的世界里，在 Service Cloud 方面，每个席位的平均价格约为 275 美元。一旦你把所有的东西都装进去，225 美元到 275 美元。
在 Agentforce 之前的世界里，在 Service Cloud 方面，每个席位的平均价格约为 275 美元。一旦你把所有的东西都装进去，225 美元到 275 美元。
Tegus Client Tegus 客户端
play_arrow00:20:53
Per month? 每月？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:20:53
Yes. It is extremely expensive. Hence why
fatigue is such a real thing.
是的。它非常昂贵。因此
， 疲劳是如此真实。
Tegus Client Tegus 客户端
play_arrow00:20:59
Got it. Something like $3,000 a year. Now, if you're able to successfully replace the workload of a person, whether it's one of these BPO workers or in-house workers that you're spending $3,000 a year with
, how would you frame how much you end up spending with
after you've replaced that person? Are you spending $6,000 with Agentforce or how much?
明白了。大约每年 3,000 美元。现在，如果您能够成功替换一个人的工作量，无论是您每年在 Salesforce 上花费 3,000 美元的 BPO 员工还是内部员工
之一，您将如何定义
在更换该人后您最终在 Salesforce 上花费了多少？您在 Agentforce 上花费了 6,000 美元还是花费了多少？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:21:23
Yes. This is where it gets hard because it's not an easy one-to-one comparison. For everything just to balance out, Agentforce needs to do 125 interactions per month successfully. A fully complete, about 125-150 interactions per month for it to be exactly equal to what you're already paying for a live agent. That's the benchmark. Very low, to be perfectly honest with you.
是的。这就是它变得困难的地方，因为它不是一个容易的一对一比较。为了平衡一切，Agentforce 每月需要成功进行 125 次互动。完全完成，每月大约 125-150 次互动，使其完全等于您已经为实时代理支付的费用。这就是基准。老实说，非常低。
是的。这就是它变得困难的地方，因为它不是一个容易的一对一比较。为了平衡一切，Agentforce 每月需要成功进行 125 次互动。完全完成，每月大约 125-150 次互动，使其完全等于您已经为实时代理支付的费用。这就是基准。老实说，非常低。
What you're seeing, the reality is once you throttle this bad boy up, that interactional volume is next to nothing of what it can actually compute. The reality is that this isn't a one-for-one exchange. You're actually seeing Agentforce is generating significantly more of what you can do there. Let's say, for instance, you would normally assume Agentforce should be, if you put into it, let's say, I'm just going to give you an example, 20 interactions per day at seven days per week times 4.3. Let's round it to 600.
你所看到的，现实是，一旦你限制了这个坏男孩，那个交互量几乎就没有它实际可以计算的东西了。现实情况是，这不是一对一的交换。您实际上看到 Agentforce 正在产生更多您可以在那里做的事情。例如，您通常会假设 Agentforce 应该是，如果您放入其中，我只举一个例子，每周 7 天，每天 20 次交互乘以 4.3。让我们将其四舍五入为 600。
你所看到的，现实是，一旦你限制了这个坏男孩，那个交互量几乎就没有它实际可以计算的东西了。现实情况是，这不是一对一的交换。您实际上看到 Agentforce 正在产生更多您可以在那里做的事情。例如，您通常会假设 Agentforce 应该是，如果您放入其中，我只举一个例子，每周 7 天，每天 20 次交互乘以 4.3。让我们将其四舍五入为 600。
At 600 engagements per month, even if I'm spending $2, that's the fully loaded, that's $1,200 per month. That's $14,400. Now, understanding those 600 engagements, if I were going to do that over with an agent, that's 5.5 minutes. 600 x 5.5 comes out to 3,300 minutes. Divide that out by 60 just so I get the hours, that's 55 hours. The thing is that's not fully accurate. That's only at 70% utilization. I'd have to do some more math on this.
每月 600 次参与，即使我花费 2 美元，这也是满载的，即每月 1,200 美元。那是 14,400 美元。现在，了解了这 600 个项目，如果我要与代理一起重新完成，那就是 5.5 分钟。600 x 5.5 相当于 3,300 分钟。把它除以 60 就这样我得到小时数，也就是 55 小时。问题是这并不完全准确。这只是 70% 的利用率。我得再做一些数学计算。
每月 600 次参与，即使我花费 2 美元，这也是满载的，即每月 1,200 美元。那是 14,400 美元。现在，了解了这 600 个项目，如果我要与代理一起重新完成，那就是 5.5 分钟。600 x 5.5 相当于 3,300 分钟。把它除以 60 就这样我得到小时数，也就是 55 小时。问题是这并不完全准确。这只是 70% 的利用率。我得再做一些数学计算。
My point is, I know for a fact it's about 30-ish percent savings right now that I would actually get of a direct labor to productional cost, even at the full $2. We know it's just going to get better as the system gets more tuned in. What you're going to find is customers will spend probably about 2X the amount of money, but they're also going to be servicing 3X-4X the amount of actual traffic.
我的观点是，我知道现在大约节省了 30% 左右，即使以 2 美元的价格，我实际上也可以从直接的劳动力到生产成本中节省。我们知道，随着系统的调整越来越多，情况会变得更好。您将发现，客户可能会花费大约 2 倍的钱，但他们也将提供 3 倍到 4 倍的实际流量。
我的观点是，我知道现在大约节省了 30% 左右，即使以 2 美元的价格，我实际上也可以从直接的劳动力到生产成本中节省。我们知道，随着系统的调整越来越多，情况会变得更好。您将发现，客户可能会花费大约 2 倍的钱，但他们也将提供 3 倍到 4 倍的实际流量。
This is where it gets really difficult to answer this question of what can we assume. The huge X factor in the room here is because this isn't just like every customer can turn it on and get these results. The X factor is what's the quality of your data inside of
? Are you using Data Cloud? Do you have essential data? Have you actually been good about structuring your data and your overall interactions inside of
today? If any of those answers to any of those questions are no, your results are exponentially going to be reduced fast.
这就是回答我们可以假设什么的问题真正困难的地方。这里房间里的巨大 X 因素是因为这不仅仅是每个客户都可以打开它并获得这些结果。X 因素是 Salesforce 内部的数据质量如何
？您在使用 Data Cloud 吗？您有重要数据吗？今天，您真的在 Salesforce 内部构建数据和整体交互方面做得好吗
？如果这些问题中的任何一个答案是否定的，那么您的结果将呈指数级快速减少。
Tegus Client Tegus 客户端
play_arrow00:23:53
Okay. Let's go into that more. The customers you're working with on those pilots, do they all use Data Cloud and do they have modern data lakes?
好。让我们更深入地讨论这个问题。与您合作进行这些试点的客户，他们是否都使用 Data Cloud，他们是否拥有现代数据湖？
好。让我们更深入地讨论这个问题。与您合作进行这些试点的客户，他们是否都使用 Data Cloud，他们是否拥有现代数据湖？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:24:02
All of them do have data lakes except for one, but not all of them are with Data Cloud. I'd say half are with Data Cloud, half are not. The half who are not now are starting to entertain utilization of it though. We've already shown them too, and we were unfortunately able to see it, that it just makes sense to do it just to increase ROI. Again, remember, all of these systems, these agent systems are algorithms.
除了一个之外，它们都确实有数据湖，但并非所有数据湖都使用 Data Cloud。我想说的是，一半使用 Data Cloud，一半没有。不过，现在没有的一半人开始享受它。我们也已经向他们展示了，不幸的是，我们能够看到，仅仅为了提高投资回报率而这样做是有意义的。同样，请记住，所有这些系统，这些代理系统都是算法。
除了一个之外，它们都确实有数据湖，但并非所有数据湖都使用 Data Cloud。我想说的是，一半使用 Data Cloud，一半没有。不过，现在没有的一半人开始享受它。我们也已经向他们展示了，不幸的是，我们能够看到，仅仅为了提高投资回报率而这样做是有意义的。同样，请记住，所有这些系统，这些代理系统都是算法。
These algorithms are only as powerful as the data you present to them to mature and to be able to educate them. To do that, you need structured data. That's where Data Cloud comes in. Again, it's next to it. It's next to that algorithm and that algorithm knows those tables and it knows its formatting. It can actually work through it better and more efficiently. I'm not having to educate the algorithm on the data and its formatting already.
这些算法的强大程度取决于您提供给他们的数据才能成熟并能够教育他们。为此，您需要结构化数据。这就是 Data Cloud 的用武之地。同样，它就在它旁边。它位于该算法旁边，该算法知道这些表格，并且知道其格式。它实际上可以更好、更高效地完成它。我不必对算法进行数据及其格式方面的培训。
这些算法的强大程度取决于您提供给他们的数据才能成熟并能够教育他们。为此，您需要结构化数据。这就是 Data Cloud 的用武之地。同样，它就在它旁边。它位于该算法旁边，该算法知道这些表格，并且知道其格式。它实际上可以更好、更高效地完成它。我不必对算法进行数据及其格式方面的培训。
Tegus Client Tegus 客户端
play_arrow00:24:54
If I make the assumption that most
customers haven't really modernized their data yet and aren't using Data Cloud yet, is that becoming a barrier to Agentforce adoption and it's slowing down the progress, or are people so committed to Agentforce that it's really just accelerating Data Cloud adoption and they're committed to modernizing their data so they can then take advantage of Agentforce?
如果我假设大多数
客户还没有真正实现数据现代化，也没有使用 Data Cloud，这是否会成为 Agentforce 采用的障碍并减慢进度，或者人们是否如此致力于 Agentforce，以至于它实际上只是加速了 Data Cloud 的采用，他们致力于实现数据现代化，以便他们能够利用 Agentforce？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:25:19
What I would tell you is most companies do not even know that is a thing yet. Again, we are very blessed in the fact that we had some very early on customers. We are seeing things that most people do not even know is there yet. That's the reality of it, is that most of these companies who are coming on Agentforce, they haven't gotten that far into the weeds yet to see that Data Cloud is going to be something that they are going to end up needing to go with. I will tell you that it is. It helps improve performance significantly.
我想告诉你的是，大多数公司甚至还不知道这是一回事。同样，我们非常幸运，因为我们有一些非常早期的客户。我们看到的是大多数人甚至不知道的东西。这就是现实情况，大多数开始使用 Agentforce 的公司，他们还没有深入研究 Data Cloud，他们最终需要使用 Data Cloud。我会告诉你，是的。它有助于显著提高性能。
我想告诉你的是，大多数公司甚至还不知道这是一回事。同样，我们非常幸运，因为我们有一些非常早期的客户。我们看到的是大多数人甚至不知道的东西。这就是现实情况，大多数开始使用 Agentforce 的公司，他们还没有深入研究 Data Cloud，他们最终需要使用 Data Cloud。我会告诉你，是的。它有助于显著提高性能。
is upfront with it. They do say Data Cloud does help and it helps more than they want to say it does. We even have customers on Snowflake instances that we set up and everything like that. That has helped a lot too. We are pointing them at an external system, an export data set for their data lake. I think what you are going to end up finding is that as this system rolls out more and more, my take on is that you are going to see companies adopting Data Cloud to be able to get that next level from these system from Agentforce, because they will start hitting barriers.
对此很坦率。他们确实说 Data Cloud 确实有帮助，而且它的帮助比他们想说的要大。我们甚至在我们设置的 Snowflake 实例上也有客户，诸如此类。这也帮了大忙。我们将他们指向一个外部系统，即他们数据湖的导出数据集。我认为您最终会发现，随着该系统越来越多地推出，我的看法是，您将看到公司采用 Data Cloud，以便能够从 Agentforce 的这些系统中获得下一个级别，因为他们将开始遇到障碍。
Now, will that be a deterrent and a slowdown? Absolutely, because Data Cloud is not cheap. Also too, because companies are sick of paying
money. With that being said, this is one of the very few
tools that I've seen personally roll out where I could actually sit here and say, "This is actually working." It's not like all the other tools that
rolls out and it's BS for at least a couple years until they actually figure it out. This tool is actually doing something.
现在，这会是一种威慑和减缓吗？当然，因为 Data Cloud 并不便宜。还有，因为公司厌倦了向
支付费用。话虽如此，这是我亲自看到的为数不多的
工具之一，我可以坐在这里说：“这确实有效。它与
推出的所有其他工具不同，在他们真正弄清楚之前，它至少需要几年的时间。这个工具实际上在做一些事情。
Tegus Client Tegus 客户端
play_arrow00:26:57
Do you think there are customers that can get a lot of value out of Agentforce without Data Cloud, or do you hit a wall pretty quickly?
您认为有些客户可以在没有 Data Cloud 的情况下从 Agentforce 中获得大量价值，还是您很快就碰壁了？
您认为有些客户可以在没有 Data Cloud 的情况下从 Agentforce 中获得大量价值，还是您很快就碰壁了？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:27:05
No. Especially your smaller businesses. Your SMB and mid-market can get an enormous amount of ROI from it because they're not in the sophisticated aspects of the actual systems yet. A lot of the actual use case and need of agent systems right now are very rudimentary and low level. It's bigger companies, more matured companies where the complexity is more there because of themselves than it is even because of their customer.
不。尤其是您的小型企业。您的 SMB 和中端市场可以从中获得巨大的投资回报率，因为他们还没有进入实际系统的复杂方面。目前代理系统的很多实际用例和需求都非常初级和低级。这是更大的公司，更成熟的公司，他们自己的复杂性比客户更复杂。
不。尤其是您的小型企业。您的 SMB 和中端市场可以从中获得巨大的投资回报率，因为他们还没有进入实际系统的复杂方面。目前代理系统的很多实际用例和需求都非常初级和低级。这是更大的公司，更成熟的公司，他们自己的复杂性比客户更复杂。
That's what a lot of the working through is actually having to be done. Without Data Cloud, you could still do it as long as your Salesforce instance has your transactional data inside of it and that you're actually working within it. Now, on the contrast to this, the issue that does occur though, it's not about size of your business, it's about the maturity of the company.
这就是许多工作实际上必须完成的工作。如果没有 Data Cloud，只要您的 Salesforce 实例中包含您的事务数据并且您确实在其中工作，您仍然可以这样做。现在，与此相反，确实出现的问题不是你的业务规模，而是公司的成熟度。
这就是许多工作实际上必须完成的工作。如果没有 Data Cloud，只要您的 Salesforce 实例中包含您的事务数据并且您确实在其中工作，您仍然可以这样做。现在，与此相反，确实出现的问题不是你的业务规模，而是公司的成熟度。
The more immature the business is, so the younger the business is, the problem that they have is the less best practices that they have inside their business which results in less data inside your CRM and less continuity of that data inside of your CRM. That is where you start running into issues. Even if I applied Data Cloud into your instance, my problem is going to be is that the quality of the data was inside of your system is not there, and/or the best practice rules or enforcement of your team inputting that data set into your system probably isn't there.
企业越不成熟，所以企业越年轻，他们的问题是他们在企业内部的最佳实践就越少，这导致 CRM 中的数据减少，CRM 内部数据的连续性降低。这就是您开始遇到问题的地方。即使我将 Data Cloud 应用到您的实例中，我的问题是您的系统内部的数据质量不存在，并且/或者您的团队将数据集输入到您的系统中的最佳实践规则或执行可能不存在。
企业越不成熟，所以企业越年轻，他们的问题是他们在企业内部的最佳实践就越少，这导致 CRM 中的数据减少，CRM 内部数据的连续性降低。这就是您开始遇到问题的地方。即使我将 Data Cloud 应用到您的实例中，我的问题是您的系统内部的数据质量不存在，并且/或者您的团队将数据集输入到您的系统中的最佳实践规则或执行可能不存在。
Tegus Client Tegus 客户端
play_arrow00:28:30
Got it. I hear a lot of stuff about 70% of companies, 80% of large enterprises still haven't cleaned up their data and don't have it organized well in a data lake. Does that seem a fair way to bucket the group? Would that mean 70% or 80% can't really do Agentforce implementations at scale, or would you size it differently?
明白了。我听到很多关于70%的公司，80%的大型企业仍然没有清理他们的数据，也没有在数据湖中很好地组织数据。这似乎是一个公平的分组方式吗？这是否意味着 70% 或 80% 的人无法真正大规模地进行 Agentforce 实施，或者您会以不同的方式调整其规模？
明白了。我听到很多关于70%的公司，80%的大型企业仍然没有清理他们的数据，也没有在数据湖中很好地组织数据。这似乎是一个公平的分组方式吗？这是否意味着 70% 或 80% 的人无法真正大规模地进行 Agentforce 实施，或者您会以不同的方式调整其规模？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:28:51
I think your number is right on the mark. Just about 70% of businesses definitely do not have data lakes or data continuity there. I think that's a correct number. The thing is though is the difference where you have to then start looking at that is you have to dig into that number then to say, "But do they have the data?" What I would tell you is that inside of that, we'll just say 70%, I would tell you probably about 80% of them have the data.
我认为你的数字是正确的。只有大约 70% 的企业肯定没有数据湖或数据连续性。我认为这是一个正确的数字。但问题是，你必须开始查看的不同之处在于，你必须深入研究这个数字，然后说，“但他们有数据吗？我想告诉你的是，在其中，我们只会说 70%，我会告诉你，他们中可能有大约 80% 有数据。
我认为你的数字是正确的。只有大约 70% 的企业肯定没有数据湖或数据连续性。我认为这是一个正确的数字。但问题是，你必须开始查看的不同之处在于，你必须深入研究这个数字，然后说，“但他们有数据吗？我想告诉你的是，在其中，我们只会说 70%，我会告诉你，他们中可能有大约 80% 有数据。
It's just not standardized, stabilized, and in a contingent fashion, but they have it. The other 20%, they have no idea. They don't even know where their data is. That 80%, it doesn't mean they can't use the system. What it means is that the time to ROI will just take longer because the system is going to have to go and try to figure out where the data is, what the data is, and how they should interpret the data. That takes a lot of time and effort.
它只是没有标准化、稳定和偶然的方式，但他们有。剩下的 20%，他们不知道。他们甚至不知道自己的数据在哪里。那 80% 并不意味着他们不能使用该系统。这意味着实现 ROI 的时间将花费更长的时间，因为系统将不得不尝试弄清楚数据在哪里，数据是什么，以及他们应该如何解释数据。这需要大量的时间和精力。
它只是没有标准化、稳定和偶然的方式，但他们有。剩下的 20%，他们不知道。他们甚至不知道自己的数据在哪里。那 80% 并不意味着他们不能使用该系统。这意味着实现 ROI 的时间将花费更长的时间，因为系统将不得不尝试弄清楚数据在哪里，数据是什么，以及他们应该如何解释数据。这需要大量的时间和精力。
Tegus Client Tegus 客户端
play_arrow00:29:48
Is this issue causing any of your pilots to slow down or take longer?
此问题是否导致您的任何飞行员放慢速度或花费更长的时间？
此问题是否导致您的任何飞行员放慢速度或花费更长的时间？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:29:52
Not yet, because it's still too new. It will though. It's definitely going to.
目前还不行，因为它仍然太新了。不过它会的。肯定会的。
目前还不行，因为它仍然太新了。不过它会的。肯定会的。
Tegus Client Tegus 客户端
play_arrow00:29:57
Okay. Let's go back to the beginning. Maybe could you just walk me through a normal timeline for how these proof of concepts work and how long does that take and how does that turn into an implementation and then how does that scale to more use cases? Walk me through the timeline for, if you can call it an average Agentforce implementation.
好。让我们回到开头。也许您能向我介绍一下这些概念验证的工作原理、需要多长时间、如何转化为实现以及如何扩展到更多用例的正常时间表吗？请向我介绍一下时间表，如果您可以将其称为普通的 Agentforce 实施。
好。让我们回到开头。也许您能向我介绍一下这些概念验证的工作原理、需要多长时间、如何转化为实现以及如何扩展到更多用例的正常时间表吗？请向我介绍一下时间表，如果您可以将其称为普通的 Agentforce 实施。
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:30:18
The good thing is with Agentforce implementations is, and this isn't us who determine this, it's just because the group who's managing all this stuff, it's all being done just like a call center implementation with. Think CCAs, UCAs, CRM, ITSM, any of these systems. They're viewing virtual agents the exact same way. What that means is POCs are ran for a minimum of six months to a year without pretty much any exception.
Agentforce 实施的好处是，这不是我们决定的，这只是因为管理所有这些东西的团队，这一切都像呼叫中心实施一样完成。想想 CCA、UCA、CRM、ITSM，这些系统中的任何一个。他们以完全相同的方式查看虚拟代理。这意味着 POC 至少运行六个月到一年，几乎没有任何例外。
Agentforce 实施的好处是，这不是我们决定的，这只是因为管理所有这些东西的团队，这一切都像呼叫中心实施一样完成。想想 CCA、UCA、CRM、ITSM，这些系统中的任何一个。他们以完全相同的方式查看虚拟代理。这意味着 POC 至少运行六个月到一年，几乎没有任何例外。
Those POCs are typically very restrictive in their scope and scale. They're usually very limited to very specific use cases and interactional types, I should say, in their scale. After six months to a year, the benchmark for conversion is actually a lot lower than you might think. It depends on the business and depends on the size of the company. I've always used what I used to call the rule of 5%.
这些 POC 的范围和规模通常非常严格。它们通常非常局限于非常具体的用例和交互类型，我应该说，就其规模而言。六个月到一年后，转化率的基准实际上比您想象的要低得多。这取决于企业和公司的规模。我一直使用我以前所说的 5% 规则。
这些 POC 的范围和规模通常非常严格。它们通常非常局限于非常具体的用例和交互类型，我应该说，就其规模而言。六个月到一年后，转化率的基准实际上比您想象的要低得多。这取决于企业和公司的规模。我一直使用我以前所说的 5% 规则。
If you can show me at least 5% of call deflection or 5% of labor arbitraging, or 5% of labor reductioning, your system will be available to go GA, general acceptance, on its current use case. I will then explore scope expansion with you. If you can get above 10%, I will immediately be exploring scope expansion and applying scope expansion at time of GA. That's my best practice. I tell you, a lot of call center and support team leadership follow something similar. They have their own methodology, but we all have our own system.
如果您能向我展示至少 5% 的呼叫转移或 5% 的人工套利，或 5% 的人工减少，您的系统就可以在其当前用例中正式发布，即普遍接受。然后，我将与您一起探索范围扩展。如果你能达到 10% 以上，我将立即探索范围扩展并在 GA 时应用范围扩展。这是我最好的做法。我告诉你，很多呼叫中心和支持团队的领导都遵循类似的东西。他们有自己的方法，但我们都有自己的系统。
如果您能向我展示至少 5% 的呼叫转移或 5% 的人工套利，或 5% 的人工减少，您的系统就可以在其当前用例中正式发布，即普遍接受。然后，我将与您一起探索范围扩展。如果你能达到 10% 以上，我将立即探索范围扩展并在 GA 时应用范围扩展。这是我最好的做法。我告诉你，很多呼叫中心和支持团队的领导都遵循类似的东西。他们有自己的方法，但我们都有自己的系统。
Tegus Client Tegus 客户端
play_arrow00:31:42
That's 5%-10% cost savings that you can identify or headcount reduction that you can identify. Is that what you're describing?
这是您可以确定的 5%-10% 的成本节省或您可以确定的裁员。这就是你所描述的吗？
这是您可以确定的 5%-10% 的成本节省或您可以确定的裁员。这就是你所描述的吗？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:31:50
That's exactly right. At 5%, just to understand, at one point, for instance, when I was helping oversee a team, it was 27,000 agents. If you reduce 5%, that's an enormous amount of labor cost savings.
这是完全正确的。5% 只是为了理解，例如，当我帮助监督一个团队时，有一次是 27,000 名代理。如果您减少 5%，则可以节省大量劳动力成本。
这是完全正确的。5% 只是为了理解，例如，当我帮助监督一个团队时，有一次是 27,000 名代理。如果您减少 5%，则可以节省大量劳动力成本。
Tegus Client Tegus 客户端
play_arrow00:32:03
You're saying if it's only getting 3%, then it's viewed as not worth your time.
你是说如果它只得到 3%，那么它就被认为不值得你花时间。
你是说如果它只得到 3%，那么它就被认为不值得你花时间。
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:32:08
I may still go GA with it, but I might expand that POC. 2%-3% is still a lot of savings. I'm going to have to look at the additional impacts of cost support administering it. I'll also probably be doing an RFP to see is there other options that are out there that are better. Could I have achieved this through my CCAs or my UCAs instead of actually having to go through my CRM and gotten the same thing? Something like that.
我可能仍然会使用它，但我可能会扩展该 POC。2%-3% 仍然是一大笔储蓄。我将不得不研究管理它的成本支持的额外影响。我可能还会做一个 RFP，看看是否有其他更好的选择。我可以通过我的 CCA 或 UCA 来实现这一点，而不是真正通过我的 CRM 并获得同样的东西吗？差不多。
我可能仍然会使用它，但我可能会扩展该 POC。2%-3% 仍然是一大笔储蓄。我将不得不研究管理它的成本支持的额外影响。我可能还会做一个 RFP，看看是否有其他更好的选择。我可以通过我的 CCA 或 UCA 来实现这一点，而不是真正通过我的 CRM 并获得同样的东西吗？差不多。
Once you get GA, then typically the scope expansions are done pretty strategically and they're usually done in quarterly or monthly implementation sprints. That's only done in Q1 and Q2. Rarely, you'll see some done in Q3 very rarely, but nothing happens in Q4. If you look at innovation in anything in support, all innovation is done in Q1, Q2, and maybe the first month, and rarely the second month in Q3, but very rarely. Q4 is locked down. Nothing happens in Q4. That just cycles for two to three years. It's all done over about a five-year period. Everything in support is done in, you look at a five-year window. That's how we work.
一旦你获得 GA，那么范围扩展通常会非常战略性地完成，它们通常在季度或每月的实施 sprint 中完成。这仅在第一季度和第二季度完成。很少，您会在第 3 季度看到一些完成，但第 4 季度没有任何变化。如果你看一下任何支持方面的创新，所有创新都在第一季度、第二季度完成，也许在第一个月完成，很少在第三季度的第二个月完成，但非常罕见。Q4 已锁定。第 4 季度没有任何变化。这只是两到三年的周期。这一切都是在大约五年的时间内完成的。所有支持工作都在 5 年窗口内完成。这就是我们的工作方式。
一旦你获得 GA，那么范围扩展通常会非常战略性地完成，它们通常在季度或每月的实施 sprint 中完成。这仅在第一季度和第二季度完成。很少，您会在第 3 季度看到一些完成，但第 4 季度没有任何变化。如果你看一下任何支持方面的创新，所有创新都在第一季度、第二季度完成，也许在第一个月完成，很少在第三季度的第二个月完成，但非常罕见。Q4 已锁定。第 4 季度没有任何变化。这只是两到三年的周期。这一切都是在大约五年的时间内完成的。所有支持工作都在 5 年窗口内完成。这就是我们的工作方式。
Tegus Client Tegus 客户端
play_arrow00:33:15
To make sure I understand, customers today, you're 18, are saying, "Let's start proof of concepts. We've identified some pretty narrow use cases of this type of customer service situation. Let's see if we can automate that and use the agent for that." You are testing that for six to 12 months and after that six to 12 months, if you believe that at scale you can get a 5%-10% headcount reduction, then you're beginning the two to three year process of achieving that 5%-10% headcount savings.
为了确保我理解，今天的客户，你 18 岁了，会说，“让我们开始概念验证吧。我们已经确定了一些非常狭窄的此类客户服务情况的使用案例。让我们看看我们是否可以自动化并使用代理来实现此目的。您将对其进行 6 到 12 个月的测试，然后在 6 到 12 个月之后，如果您相信大规模可以减少 5%-10% 的员工人数，那么您将开始为期 2 到 3 年的过程，以实现 5%-10% 的员工人数节省。
为了确保我理解，今天的客户，你 18 岁了，会说，“让我们开始概念验证吧。我们已经确定了一些非常狭窄的此类客户服务情况的使用案例。让我们看看我们是否可以自动化并使用代理来实现此目的。您将对其进行 6 到 12 个月的测试，然后在 6 到 12 个月之后，如果您相信大规模可以减少 5%-10% 的员工人数，那么您将开始为期 2 到 3 年的过程，以实现 5%-10% 的员工人数节省。
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:33:48
No. The 5% or 10% savings is done in the POC. At least that's what I'm looking for. Now, mind you, some companies, like I said, it might be less. It just depends on the business and the scale.
不。5% 或 10% 的节省是在 POC 中完成的。至少这就是我正在寻找的。现在，请注意，就像我说的，一些公司可能会更少。这只取决于业务和规模。
不。5% 或 10% 的节省是在 POC 中完成的。至少这就是我正在寻找的。现在，请注意，就像我说的，一些公司可能会更少。这只取决于业务和规模。
Tegus Client Tegus 客户端
play_arrow00:34:00
The two to three years is you're aiming for something much bigger.
两到三年是你的目标，目标要大得多。
两到三年是你的目标，目标要大得多。
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:34:04
Yes. Significantly bigger. 30%-50% kind of thing. That's why it's a three- to five-year plan. These are huge initiatives at that point. Something like Agentforce, that's the promise. Just to understand, that's the thing with these AI agents. What people are working to achieve is the benefits you've already started. I'm not saying everyone gets it because it's not everyone. What you're starting to see coming out of that BDR space, and that's why I mentioned it to you, the business development rep space, we've seen those case studies already come out. Customer service is looking for those type of success stories.
是的。大很多。30%-50% 之类的东西。这就是为什么它是一个三到五年的计划。在这一点上，这些都是巨大的举措。就像 Agentforce 这样的东西，这就是承诺。只是为了理解，这就是这些 AI 代理的问题。人们正在努力实现的是你已经开始的好处。我并不是说每个人都能理解，因为不是每个人都能理解。您开始看到 BDR 领域出现的情况，这就是我向您提到的原因，业务开发代表领域，我们已经看到这些案例研究已经出现。客户服务正在寻找这种类型的成功案例。
是的。大很多。30%-50% 之类的东西。这就是为什么它是一个三到五年的计划。在这一点上，这些都是巨大的举措。就像 Agentforce 这样的东西，这就是承诺。只是为了理解，这就是这些 AI 代理的问题。人们正在努力实现的是你已经开始的好处。我并不是说每个人都能理解，因为不是每个人都能理解。您开始看到 BDR 领域出现的情况，这就是我向您提到的原因，业务开发代表领域，我们已经看到这些案例研究已经出现。客户服务正在寻找这种类型的成功案例。
Tegus Client Tegus 客户端
play_arrow00:34:42
Okay. An average three- to five-year goal, you said 30%-50% savings?
好。平均三到五年的目标，您说的是节省 30%-50%？
好。平均三到五年的目标，您说的是节省 30%-50%？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:34:47
I'd say about 30%. 30% is a big number already. 30%, I think over the next five years, and it should be achievable. If you look at the innovational curve and how it's planing right now, we should easily be able to get to about a 30% net reduction of labor in those support teams within the next five years.
我会说大约 30%。30% 已经是一个很大的数字了。30%，我认为在未来五年内，这应该是可以实现的。如果您查看创新曲线及其目前的规划情况，我们应该能够在未来五年内轻松实现这些支持团队的劳动力净减少约 30%。
我会说大约 30%。30% 已经是一个很大的数字了。30%，我认为在未来五年内，这应该是可以实现的。如果您查看创新曲线及其目前的规划情况，我们应该能够在未来五年内轻松实现这些支持团队的劳动力净减少约 30%。
Tegus Client Tegus 客户端
play_arrow00:35:07
The six- to 12-month proof of concept, what does that look like in terms of actual spending with
? How do these contracts work? If you're doing a proof of concept, trying to prove something out over six to 12 months and then you have a bigger goal, how does a contract with
work for that whole process?
6 到 12 个月的概念验证，就 Salesforce 的实际支出而言，情况如何
？这些合同如何运作？如果你正在进行概念验证，试图在 6 到 12 个月内证明一些东西，然后你有一个更大的目标，那么与 Salesforce 的合同如何
在整个过程中发挥作用？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:35:27
Because it's consumption-based, it's very easy.
因为它是基于消费的，所以非常简单。
因为它是基于消费的，所以非常简单。
Tegus Client Tegus 客户端
play_arrow00:35:30
You're paying consumption during that six to 12-month period.
您在这 6 到 12 个月内支付消费费用。
您在这 6 到 12 个月内支付消费费用。
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:35:34
Exactly. What you're also doing is you agree to contractual minimums. Because it's still early on, those contractual minimums of transactions per month, I will tell you, they are being very flexible right now. That's what's setting the rate. What sets that rate is those contractual minimums.
完全。您还要做的是您同意合同的最低要求。因为现在还为时过早，所以那些每月的合同最低交易额，我会告诉你，他们现在非常灵活。这就是设定费率的原因。设定该比率的是那些合同最低限度。
完全。您还要做的是您同意合同的最低要求。因为现在还为时过早，所以那些每月的合同最低交易额，我会告诉你，他们现在非常灵活。这就是设定费率的原因。设定该比率的是那些合同最低限度。
Tegus Client Tegus 客户端
play_arrow00:35:52
Our customers end up paying much in that six- to 12-month period?
我们的客户最终在 6 到 12 个月的时间内支付了很多费用？
我们的客户最终在 6 到 12 个月的时间内支付了很多费用？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:35:56
Yes. If you want the biggest savings, you're agreeing to really tens and hundreds of thousands of transactions a month. The beauty of it, especially in support, we already know those numbers. That's easy. We know exactly how much volume we get.
是的。如果您想要节省最大的费用，那么您就同意每月进行数万笔和数十万笔交易。它的美妙之处，尤其是在支持方面，我们已经知道这些数字。这很简单。我们确切地知道我们获得了多少交易量。
是的。如果您想要节省最大的费用，那么您就同意每月进行数万笔和数十万笔交易。它的美妙之处，尤其是在支持方面，我们已经知道这些数字。这很简单。我们确切地知道我们获得了多少交易量。
Tegus Client Tegus 客户端
play_arrow00:36:11
You mentioned an example earlier where a customer already maybe had line of sight to half of their spending with
being on the agent side versus the human seat side. Walk me through that example. How do you see your customers spend with
evolving over the six- to 12-month period and then over the three- to five-year period? Are customers spending a lot more with
even in the six- to 12-month period if it's all working or only in the three- to five-year period do they start to spend more with
?
您之前提到了一个例子，客户可能已经可以看到他们一半的支出，因为
的支出是在座席端，而不是人工座席端。请给我看个例子。您如何看待客户在
上的支出在 6 到 12 个月以及 3 到 5 年期间的变化？如果
一切正常，客户是否在 6 到 12 个月内在 上花费更多，或者仅在 3 到 5 年内，他们才开始在
上花费更多？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:36:45
Yes. Actually, it's a pretty cool model in somewhat
's favor. Over a POC period, there's a lot of replicated costs. That's why POCs are actually so expensive because a lot of people, they don't understand is that just because we're reducing stuff and just because we're seeing tangible positive results, not at all, but most businesses do not start making cuts. You're not going to just start slashing your labor.
是的。实际上，这是一个非常酷的模型，在某种程度
上对 有利。在 POC 期间，会有大量的复制成本。这就是为什么 POC 实际上如此昂贵的原因，因为很多人不明白，仅仅因为我们正在减少开支，仅仅因为我们看到了切实的积极成果，根本不是，但大多数企业并没有开始削减开支。你不会就这样开始削减你的劳动力。
The reason is because, one, most people don't trust the data when it first starts coming out. There's a lot of questioning like, "Is this real? Are the customers actually happy? Are we going to get a flood of complaints coming in the next two to three months? What's the outcome of this?" You got that. There's also probably a ton of work that had to be done that's been being put off forever that also needs to get done.
原因是，第一，当数据刚开始出现时，大多数人不信任数据。有很多问题，比如，“这是真的吗？客户真的满意吗？在接下来的两到三个月内，我们会收到大量投诉吗？结果如何？你明白了。可能还有大量必须完成的工作被永远推迟了，也需要完成。
原因是，第一，当数据刚开始出现时，大多数人不信任数据。有很多问题，比如，“这是真的吗？客户真的满意吗？在接下来的两到三个月内，我们会收到大量投诉吗？结果如何？你明白了。可能还有大量必须完成的工作被永远推迟了，也需要完成。
Number three is, typically, which is more common, especially amongst the larger businesses, they're usually running POCs simultaneously. We usually have two things running at the exact same time. The bigger part for
, to answer your direct question, the licensing is still there. All we've done is just add cost basically right now where things start to change. With that being said, that's all upside for
. The next six to 12 months,
should see actually increases, not decreases. They should all be upside for them over the next six to 12 months.
第三点通常是更常见的，尤其是在大型企业中，他们通常会同时运行 POC。我们通常有两个东西同时运行。对于
来说，回答您的直接问题，更大的部分是许可仍然存在。我们所做的只是在情况开始发生变化的地方增加成本。话虽如此，这就是
的全部优势。在接下来的 6 到 12 个月里，
应该会看到真正的增长，而不是减少。在接下来的 6 到 12 个月内，它们都应该有上行空间。
After, coming out of 2025, going into 2026, what we should see there, or at least what we're planning on seeing is we will probably see a little dip early on as companies begin to start doing labor correctioning, and as those licensings go away, of course, then you'll have transactional volume start to increase.
在 2025 年之后，进入 2026 年，我们应该在那里看到的，或者至少我们计划看到的是，随着公司开始进行劳动纠正，我们可能会在早期看到一点下降，当然，随着这些许可的消失，交易量将开始增加。
在 2025 年之后，进入 2026 年，我们应该在那里看到的，或者至少我们计划看到的是，随着公司开始进行劳动纠正，我们可能会在早期看到一点下降，当然，随着这些许可的消失，交易量将开始增加。
This is going to come to where how well
do their job in sales. I'll give you both sides of this. Why I said there's going to be a dip is we know
is giving away the kitchen sink right now, especially in their pricing. Those early customers are going to be the customers who probably do labor cuts early, licensing cuts early, and they're going to be on cheap rates.
这将涉及
在销售方面的表现如何。我会给你两面。我之所以说会出现下降，是因为我们知道
现在正在放弃厨房水槽，尤其是在他们的定价方面。这些早期客户可能是那些可能提前裁员、提前削减许可的客户，而且他们的费率很便宜。
It's going to take them exponentially longer to get their transactional volume up high enough to exceed what
was originally seeing in their licensing spend. Not a ton, but it is going to still be pretty quite a bit versus the next gen of customers, which we'll probably start seeing in the next six to 12 months start coming in, which will be at a higher rate.
will see an ROI on those much faster because their rate should be higher because they're going to be able to start justifying the higher rate to those customers.
他们将需要成倍的时间才能使交易量增长到足够高的水平，以超过
最初看到的许可支出。不是很多，但与下一代客户相比，它仍然会相当多，我们可能会在未来 6 到 12 个月内开始看到这些客户开始进入，而且速度会更高。
将更快地看到这些 ROI，因为他们的费率应该更高，因为他们将能够开始向这些客户证明更高的费率是合理的。
My point is, one year, we'll see
should be positive. I think 2026, you're going to see a little bit of dip as licensing spend goes down. You'll see an increase in revenues from Agentforce, but it won't be as much because the net amount being charged is going to be too small even though they'll cover up some of that with the contractual minimums that they're going to be getting and probably doing some of that. I still think you'll see a net reduction in some revenues, net net. 2027, 2028 is where you start seeing up on it. That's how we view it. 2025 should be very good for them, in my opinion.
我的观点是，有一年，我们会看到
应该会是积极的。我认为 2026 年，随着许可支出的下降，你会看到一点下降。你会看到 Agentforce 的收入有所增加，但不会那么多，因为收取的净额会太小，即使他们会用他们将要得到的合同最低限额来掩盖其中的一些，并且可能会做一些。我仍然认为你会看到一些收入的净减少，净净收入。2027 年、2028 年是你开始看到它的时候。这就是我们的看法。在我看来，2025 年对他们来说应该是非常好的。Tegus Client Tegus 客户端
play_arrow00:40:11
If these things work, as you map out, if these six- to 12-month proof of concepts work and then you greenlight the next three to five years and those work, If you're successful in reducing 30% of your head count, does total spend with
double or triple? What do you think? How does the math end up at the end state?
如果这些事情奏效，正如你所规划的那样，如果这些 6 到 12 个月的概念验证奏效，然后你为接下来的 3 到 5 年开绿灯，这些都奏效了，如果你成功地减少了 30% 的员工人数，那么在 Salesforce 上的总支出是
两倍还是三倍？你觉得怎么样？数学如何以结束状态结束？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:40:31
This is where things get a little interesting. We believe it almost doubles based from what we see. The X factor in it, which there always is one, is how good is the business. If the business is just supporting the same amount of customers and not growing or if it starts retracting, that has a direct line impact because it's consumptional based. If you look at current trends in full interactional models, the reality is this communication demand into a business right now have been growing up anywhere from 25%-40% year-over-year for the last almost.
这就是事情变得有点有趣的地方。我们认为，根据我们所看到的情况，它几乎翻了一番。其中的 X 因素（总是有一个）是业务的好坏。如果企业只是支持相同数量的客户而没有增长，或者如果它开始收缩，那会产生直接的直线影响，因为它是基于消费的。如果你看一下完全交互模型的当前趋势，现实情况是，现在对企业的这种通信需求几乎在过去几年里每年增长 25%-40%。
这就是事情变得有点有趣的地方。我们认为，根据我们所看到的情况，它几乎翻了一番。其中的 X 因素（总是有一个）是业务的好坏。如果企业只是支持相同数量的客户而没有增长，或者如果它开始收缩，那会产生直接的直线影响，因为它是基于消费的。如果你看一下完全交互模型的当前趋势，现实情况是，现在对企业的这种通信需求几乎在过去几年里每年增长 25%-40%。
Now, mind you, that's not live call because live call has actually been diminishing year over year. Total net traffic demand into support has been increasing, like I said, exponentially, especially last five to six years. A lot during COVID. We don't see that slowing, especially with all the channels out there in the omnichannel. If you take that and you figure that into the equation,
could potentially 10X their revenues as long as it all stays consumptional along the way. I think a conservative number would be 2X-3X.
现在，请注意，这不是实时通话，因为实时通话实际上每年都在减少。正如我所说，用于支持的总净流量需求一直在呈指数级增长，尤其是过去五到六年。在 COVID 期间很多。我们没有看到这种情况会放缓，尤其是在全渠道中的所有渠道中。如果你把这个算进去，
只要 在整个过程中保持消费，他们的收入就有可能增加 10 倍。我认为保守的数字应该是 2X-3X。
Tegus Client Tegus 客户端
play_arrow00:41:38
What portion of those 18 from proof of concept, from what you're seeing, do you think have a good chance of moving forward to the next step or what portion end up getting bogged down or slowed down and have to fix something else first?
从您所看到的，这 18 个概念验证中，您认为有很大的机会进入下一步，或者哪一部分最终陷入困境或放慢速度，必须首先修复其他问题？
从您所看到的，这 18 个概念验证中，您认为有很大的机会进入下一步，或者哪一部分最终陷入困境或放慢速度，必须首先修复其他问题？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:41:53
I'd say right now, about 30%-40%. Let me put in real numbers. I'd say about six of the 18 are probably I see going long term with this. There's about four to six, so about another third of them, they're running POC simultaneously with one of their other systems. A popular one is with ServiceNow. They're running head-to-head ServiceNow and Agentforce just to see who plays out. The other third, a lot of their problems are with their data. It's not that they're not getting the ROI, it's just been slow, and so I think it's just going to take a little bit longer.
我想说现在，大约 30%-40%。让我输入真实的数字。我想说的是，这 18 个中大约有 6 个可能是我认为会长期使用的。大约有 4 到 6 个，因此大约还有三分之一，他们与其他系统同时运行 POC。一个流行的是 ServiceNow。他们与 ServiceNow 和 Agentforce 正面交锋，只是为了看看谁能发挥作用。另外三分之一，他们的很多问题都与他们的数据有关。这并不是说他们没有获得投资回报率，只是速度很慢，所以我认为这需要更长的时间。
我想说现在，大约 30%-40%。让我输入真实的数字。我想说的是，这 18 个中大约有 6 个可能是我认为会长期使用的。大约有 4 到 6 个，因此大约还有三分之一，他们与其他系统同时运行 POC。一个流行的是 ServiceNow。他们与 ServiceNow 和 Agentforce 正面交锋，只是为了看看谁能发挥作用。另外三分之一，他们的很多问题都与他们的数据有关。这并不是说他们没有获得投资回报率，只是速度很慢，所以我认为这需要更长的时间。
Tegus Client Tegus 客户端
play_arrow00:42:28
Okay. The ones where they're running ServiceNow head to head, which is looking better?
好。他们正面交锋运行 ServiceNow 的那些，哪个看起来更好？
好。他们正面交锋运行 ServiceNow 的那些，哪个看起来更好？
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
play_arrow00:42:34
ServiceNow. I think ServiceNow is better. Even in CSM, I think ServiceNow system is superior.
ServiceNow 的我认为 ServiceNow 更好。即使在 CSM 中，我认为 ServiceNow 系统也更胜一筹。
ServiceNow 的我认为 ServiceNow 更好。即使在 CSM 中，我认为 ServiceNow 系统也更胜一筹。
Tegus Client Tegus 客户端
play_arrow00:42:39
Okay. Well, thank you again for taking the time to speak with us today. This was very helpful. Enjoy the rest of your day.
好。好了，再次感谢您今天抽出时间与我们交谈。这非常有帮助。享受你一天的剩余时间。
好。好了，再次感谢您今天抽出时间与我们交谈。这非常有帮助。享受你一天的剩余时间。
done_all
Read 读
account_circle
Chief Executive Officer at BizSprints
BizSprints 首席执行官
BizSprints 首席执行官
$600
estimated rate / hr 预计速率 / 小时
10
interviews on Tegus Tegus 访谈
Other Interviews from this Expert
本位专家的其他访谈
本位专家的其他访谈
Former Exec 前任执行官
Former Account Executive - Financial Services at Salesforce
Salesforce 金融服务前客户经理
21/3/19Salesforce 金融服务前客户经理
Former Exec 前任执行官
Former Account Executive- Financial Services at Salesforce
Salesforce 金融服务部前客户经理
20/8/4Salesforce 金融服务部前客户经理
Partner 伙伴
Chief Executive Officer at BizSprints
BizSprints 首席执行官
24/12/13BizSprints 首席执行官
Customer 客户
Chief Executive Officer at BizSprints
BizSprints 首席执行官
24/11/14BizSprints 首席执行官
Partner 伙伴
Chief Executive Officer at BizSprints
BizSprints 首席执行官
24/10/7BizSprints 首席执行官
Partner 伙伴
Chief Executive Officer at BizSprints
BizSprints 首席执行官
24/5/31BizSprints 首席执行官
Partner 伙伴
Chief Executive Officer at BizSprints
BizSprints 首席执行官
24/5/21BizSprints 首席执行官
The information contained herein is for informational purposes only, is not to be construed as investment advice, and should not be used as the sole basis for making any investment decision.
Tegus, Inc. ("Tegus") makes no representations and accepts no liability for the content herein or for any errors, omissions, or inaccuracies, and will in no way be held liable for any potential or actual violations of United States laws, including without limitation any securities laws, based on the information herein. The views expressed in this document are those of the expert or the interviewer, and they are not endorsed by, nor do they represent the opinion of, Tegus. Tegus reserves all copyright, intellectual, and other property rights in this document and other Site Content, which are protected by the Copyright Laws of the United States and may not be copied, reproduced, sold, published, modified, or exploited in any way without the express written consent of Tegus.