Share
Author
Andy Martin
Share
I Run a Software Company. Your CSRs Are Paying the Price.
In my last two posts, I made two admissions.
First: I run a software company. And I still buy software. I don’t build everything. Sometimes the smartest move is to buy what already works.
Second: I run a software company. And I don’t always recommend us. Not every problem is a fit. Define the problem first. Then pick the tool.
Here’s the third.
I run a software company. And I don’t trust what my industry is doing to your customer service teams.
If you lead customer service for a manufacturer running SAP, this probably feels familiar. Reps are overwhelmed. The tech stack keeps growing. Turnover keeps rising.
The numbers are ugly. Over 60% of call center agents report burnout risk. Many B2B contact centers live in a world where annual turnover is normal, not alarming. The work is high volume. The systems are fragmented. The day is constant context switching.
At the same time, expectations keep climbing. AI. Omnichannel. Digital transformation.
SAP’s own guidance on AI in service includes a point that should not be controversial, but somehow is: AI should be anchored in real workflows, not deployed for its own sake. The fact that this warning needs to be said at all should give anyone in leadership a pause.
I’ve sat in the boardrooms where “platform expansion” is the growth lever and ARR is the headline. I’ve also noticed what is usually missing from the slide deck.
Nobody asks what a CSR’s screen looks like at 10:17 a.m. when a dealer calls and says, “Where’s my order?”
That is the question that matters.
The hidden tax on the front line
In a typical SAP centric service environment, answering one basic order status question can mean jumping across a handful of systems: SAP, CRM, Email, carrier sites, a portal that may or may not be current.
So CSRs become human APIs. They translate and reconcile information all day, then get measured on speed.
Research is consistent on the pattern: fragmented tool stacks and cognitive overload show up as longer handle times, higher error rates, and lower satisfaction. Context switching is not just annoying. It is expensive. And the cost compounds into stress, rework, and operational debt.
Vendors celebrate omnichannel. Your team experiences omni-swivel-chair.
Now layer in AI.
Chatbots. AI agents. Auto summaries. Suggested replies.
A lot of companies talk about this as relief. But many teams experience it as supervision. Someone still has to verify the output, correct it, and clean up the edge cases. Nearly 80% of workers using generative AI say it has increased their workload because they spend time reviewing and correcting AI output.
If AI does not reduce the number of systems a rep touches to resolve a ticket, it is not innovation. It is noise.
If it doesn’t eliminate a step, don’t take it
Here’s the rule I give service leaders when someone proposes a new tool.
For every system you add, tell me what step disappears.
A properly integrated SAP self-service portal should allow customers to:
Check order status directly against SAP.
Retrieve invoices without emailing a CSR.
Confirm pricing from the system of record.
Studies on well designed self service consistently show routine inquiry volume can drop dramatically, often 50 to 70 percent, when customers can reliably get answers on their own.
If the stack only grows, the burden grows.
If a project does not move first contact resolution, handle time, CSAT, or deflection, it is decoration.
And since I’m the CEO of a software company, I’ll say it plainly:
If we cannot show a believable path to improving those numbers, you should not buy from us.
When I walk away
There are times I tell a prospect no. Not because the software is bad. Because I’ve seen the same failure pattern too many times.
I walk away when there is no service leader insisting on fewer system touchpoints, and the initiative is being driven as a technology project.
I walk away when SAP master data is a mess and nobody is willing to own the cleanup. A portal can only expose what the ERP can support. Bad data does not become trustworthy just because it has a nicer UI.
I walk away when the organization is not willing to retire legacy tools and enforce a new workflow. If everything stays, nothing changes.
There are times when portals “go live” and call volume go up because customers didn’t trust the information. That outcome is worse than doing nothing. It adds cost and disappointment at the same time.
Good software is subtractive. It removes steps, reduces screens, and makes work simpler.
CSRs live with the consequences of every system decision.
Make sure the software does too.




