This article is Part Two of a three-part learning series about profitability, scale, and digital transformation in support teams at direct-to-consumer and digital retail brands. Part One is available here.
Scenario: You work at a direct-to-consumer brand. You lead the support department and help keep customers happy. You’ve just successfully stuck up for yourself and also earned recognition from other leadership team members that your department is not a cost center, but a catalyst for accelerating growth. Your team can run so efficiently that you can boost margins without compromising brand integrity.
Now they ask, “So what improvements are you working on next?”
In terms of efficiency, there’s really only one answer. You need something that will earn back time for your agents and handle tickets independently.
You need automation.
BUT WAIT! The reason why your support team is so excellent and your brand has a good reputation is because you’re more human, and you actually care. You’re not just another big logo with a conveyor-belt self-service experience. So, won’t letting the “robots” handle support tickets ruin what your team has built so far?
It all depends on how you plan to leverage automation. The key is to use it at a point where you know actions become standardized and procedural. Processing a refund, canceling an order, giving a tracking update – those are all handled differently from each other, but the answers for each of them probably look the same regardless of whether the response is automated, or if a human answered it.
Let’s look at it in terms of a very human-looking dialogue:
Now, let’s convert this interaction to the basic series of actions that took place:
Take a step back for a second and think about the customer’s goals and the brand’s goals.
The customer wants to complete their return, and check this chore off their list.
The brand wants to adhere to policy, but otherwise, delight the customer and exceed expectations.
Let’s simplify further:
Automation thrives in situations that involve simplicity and following precise instructions, and it’s always going to be faster than a human. So this is pretty perfect. Well, almost.
The big risk factor with automation is that if the instructions they follow aren’t defined very well, the actions that follow won’t be quite right. And then nobody’s happy, because the robot has gone off the rails and is doing its own thing, the brand’s policies may be compromised, and the customer didn’t get what they want.
The safety rail, then, that helps ensure automation provides the efficiency and profitability benefits without introducing heavy risks is AI.
The radical swings in quality between a fully automated system and a human-driven one usually from what happens at the very start of a process. Humans are very good at reading subtext and context and can interpret the intended outcome. And usually, machines aren’t good at that.
But with the right AI model that leverages natural language understanding (NLU), the automation can handle that interpretation action quite well, especially in well-defined cases that it’s trained for.
Now with AI, instead of a risky rampage bot that’s brutally efficient, you have a non-risky friendly bot that’s brutally efficient. But in the way you want.
The Low Down on Labor
Let’s say you have a team of three support agents, each working at 100% capacity. They’re tapped out. Your customer base continues to grow, though, which means the support ticket workload per agent is increasing.
Normally, when you reach resource limits, you need to invest in more of that resource; in this case, hiring at least one additional agent. Let’s say the additional hire costs $60,000 all-in. Let’s also say that that agent will take about six months of ramp time before they also are at 100% capacity.
On a regular pace for growth, that means you’re investing at least $120,000 per year on labor, plus the additional costs of recruiting, training, and other peripheral investments associated with the hiring process.
Contrast this with the alternate approach, which would be to make your agents more efficient. Let’s say you implement your shiny new AI and automation support platform for the same $60,000, but this time it’s for the full year.
On day one, your support platform starts working by automatically resolving some portion of your support tickets and eliminating that set of work from your agents. On top of that, the platform provides additional value by helping your agents work faster by surfacing data from other integrated tools, like your order management system (OMS).
Support agent capacity, on day one, goes from 100% across three agents to 75% across three agents. And if we already know it’s about six months for an agent to hit 100% capacity, the brand just bought itself another four months or so before another agent has to be hired.
Furthermore, the presence of the automation tool lengthens the time-to-capacity from six months to twelve months because of how much volume it picks up.
That’s a little complicated, so let’s do a side-by-side comparison of investment pacing for each model:
Model A, Continually Hire Support Agents:
- $120,000+ (2) agent hires per year
- Reps are constantly at capacity
- Status quo in customer experience with support
Model B, Supercharge Support Agents with Platform:
- $60,000 platform cost per year
- $60,000+ (1) agent hire per year
- Reps at comfortable capacity
- Customer tickets are engaged and resolved more quickly
At a minimum in this scenario, we’re looking at a reduced investment in hiring and an improvement in customer experience with Model B.
And that’s just a mock example. Real results from customers who adopted the Thankful AI and automation solution include:
So what does an AI automation customer support solution give back to the brand?
Longer timeframes without additional personnel cost, meaning that a whole lot more sales growth is translating to pure profits. That’s the power of profitability, and that’s the impact of optimizing support operations.