The ServiceDNA™ Diagnostic
Your best people are doing two jobs.
Find out where the second one is costing you.
The first job is the one you’re paying for. The second is everything your people do to get the first one done: rechecking work they don’t trust, untangling a promise two teams understood differently, finding the one person who can settle a decision. It’s written in invisible ink: your systems record the ticket, not the three people it took to close it, so your AI misses it, and it’s hard to see from where you sit.
You’re about to put money behind something: a new operating model as the company scales, an AI rollout, an acquisition to integrate, a team that keeps missing its number. Your dashboards show the missed target and the backlog. They don’t show whether that team is the problem or is covering for one, and that decides where the money should go.
The ServiceDNA Diagnostic asks everyone whose work touches the customer fourteen questions about getting the work done, then groups people by how they answered, whatever department they sit in. It shows which conditions run across the business, which are concentrated in particular places, and how those patterns appear along the path from what you promise customers to what you deliver. The survey can’t tell a new team from a team absorbing decisions made upstream; I follow the work to find out which. In about thirty days your leadership team knows where to investigate first, and where not to spend yet.
14 diagnostic questions · anonymous · about 5 minutes · historically over 90% response · 30 days end to end
Example from one operation. Each bar shows the share of a function in a response pattern associated with getting help quickly, trusting what comes back and being able to act: about 40% in Support, 20% in Professional Services and 16% in Client Success. The mature Support knowledge base, one-off work in Professional Services and a newer Client Success function helped explain the differences. The chart raises the question; the work behind it supplies the explanation.
When is the Diagnostic worth running?
Before a consequential investment. You’re committing money and people to a new platform, a knowledge program or a different way of delivering as the business scales. You want to see whether the people expected to make it work have what they need, and where conditions might make the change harder.
When problems keep coming back. Each team fixes its piece, and the same problem returns. It lives in the handoffs, where nobody owns it.
Before you reorganize for AI. You’re changing roles, teams or workflows to put AI to work. You want to see how the work actually moves, and who is carrying it, before you redraw the lines.
In your first months in a role. You need to see how the work actually runs before you decide what to change.
After an acquisition, or across a portfolio. The businesses you’ve brought together work under different assumptions, and you need to know where integration needs attention first. For a PE firm, every portfolio company gets the same starting read: the same fourteen questions, scored the same way. What the patterns mean is worked out inside each company.
What your existing measures can miss
A team can be highly engaged and still be doing two jobs. A strong engagement score can sit on top of a manager or veteran holding a broken process together by hand. Another engagement survey won’t tell you what to change. The Diagnostic points to conditions worth investigating: a process that makes people recheck decisions, a measure that pulls two groups in opposite directions, the one person everyone needs before anything moves. Those are things leaders can change.
Every team’s dashboard can be right while the whole operation struggles. Sales hits its target, implementation misses a date, and service absorbs the exceptions. Each number tells the truth about its own team. None explains what happened to the promise, the decision or the knowledge as the work moved between them. The same role can also be a different job across locations, even on identical systems. A company-wide rollout spreads effort like peanut butter: a thin layer everywhere, and nothing moves far enough to matter.
The Diagnostic shows where people report different conditions; following the work helps leaders see where a shared change makes sense, where attention belongs first, and which teams already have something others could learn from.
Case Study
They thought they had a knowledge problem.
The situation. A legal services firm had tried twice to fix its knowledge problem. Both programs failed, and AI was starting to look like a threat to the people who would have to use it. Before a third try, they ran the Diagnostic. More than 90% of their people answered.
What we found. Fewer than half of respondents fell into response patterns associated with being able to find and rely on knowledge. The firm was four merged businesses that had never been brought together in how they worked. The people were collegial, but the same words meant different things in each business, and they didn’t trust what came across from the others.
The firm also ran on standard operating procedures (SOPs): multi-page documents that spell out each process step by step. Finding an answer meant reading through one, and changing anything, however small, meant revising one.
Both earlier programs had tried to fix the knowledge. Neither had dealt with four businesses that didn’t trust each other’s work, or with documents nobody could change quickly.
What they did next. After the readout, they brought me in to coach a pilot team. The pilot took one piece of work that crosses all four businesses: setting up a new company for a client, which takes an attorney, a tax professional and a business advisor working in sequence.
People from each business walked through it the way the client lives it. They had sixty years at the firm between them, and they found a step none of them knew existed. So the people doing the work wrote down how it gets done, across the four businesses, and agreed what the words meant as they went.
Because every change had always waited on an SOP, we kept each one small: fix one thing, check it lightly, measure it. Six months went into building what the program needed to run: the coaches, the training, the measures of success and the technology. Then they launched.
What happened. People can now find the answer they need, trust it and use it, without reading through a multi-page SOP. Setting up a company for a client takes much less time than it did, and the ones that still run long are the exceptions.
People across the four businesses know each other’s workloads and problems, and they offer to help. Training a new person to work this way dropped from ten weeks to five as they kept improving it.
A senior attorney who had been skeptical of AI built an AI assistant that drafts knowledge articles in the client’s words, checks them against the team’s standard, and coaches the person writing. Every member of the pilot team volunteered to mentor the next team, and that team did the same. Other teams started asking to be next.
The thirty-day Diagnostic helped the firm choose where to begin. The improvements came from the sustained work of its people afterward, supported by a separate engagement.
How does the Diagnostic work?
The ServiceDNA Diagnostic is a thirty-day assessment of how work moves through your operation. It combines a fourteen-question anonymous survey, statistical grouping and AI-assisted interpretation, targeted conversations and a leadership readout.
The survey finds the patterns. I set them against how the work moves, the measures each group answers to, and the incentives behind what people do. I test the explanations with the people doing the work, their managers and your leaders. The readout turns what survives into choices about where to act.
Everyone answers the same fourteen questions in the first week
Everyone whose work touches the customer answers, anonymously, in about five minutes. The questions cover what people know, what they have to work with, and how the work treats them. The fourteen never change. Up to four more are designed with you, so you can slice the results by business unit, location, tenure, or remote and in-office. In every engagement so far, more than 90% of the people invited have answered, across thousands of responses. The fourteen questions and the statistical method behind them form the core of ServiceDNA.
Within ten days, people are grouped by how the work is for them
The answers go through the same statistical method every engagement. People are grouped by the pattern of their answers rather than by title or department, so each group represents people reporting similar working conditions. How many groups there are depends on your data.
Each group is broken out by your slices, so you see which business units it shows up in and which it doesn’t. Each also gets a name for what it’s living through, never a judgment of the people in it.
That gives teams who have had the same argument for years a way to talk about it without anyone being the villain.
The groups are laid along the path your work travels
Your org chart shows who reports to whom. It doesn’t show how the work moves: from what’s promised to the customer, to what gets delivered, to who inherits it afterward.
Laying the groups along that path shows where people working under different conditions meet, usually where the work changes hands, in places no org chart would point you to.
Those are the places to look: is knowledge stopping there, are decisions waiting, is one group covering for another? Following the work, with the people doing it and the measures they answer to, answers that.
Conversations test what’s behind the patterns
I talk with three sets of people.
First, two or three people who fit each group’s description, from the points where the work changes hands. I describe the group by how its members experience the work, you suggest people who fit, and neither of us knows how any of them answered.
Second, the managers of those groups, who live with these patterns every day.
Third, your leaders, who see what the work is supposed to add up to. I watch how the work actually runs and ask about the workarounds and the unofficial version of the work. I also go through the data, reports, measures and incentives the work runs on.
In a one-hour readout, your leaders work out what to fix first
Your leadership team sees the operation the way the work travels through it, together, often for the first time. I bring the groups, the flow and what the conversations found. You bring what I can’t know: the strategy, the history between your teams, the past attempts, the acquisitions and reorganizations. The order of what to fix first gets worked out in the room, and your leadership team leaves knowing where to start. The readout brings operating evidence and leadership context into the same conversation.
What do you leave with after thirty days?
Your leadership team leaves knowing where people report different working conditions, which explanations held up when we followed the work, and where to start.
The report lands in your hands at the end of the readout, not before: the groups, the flow, what the conversations surfaced, how it relates to your existing measures, and the sequence worked out in the room. Each business unit walks out with its own reading, and the operation walks out with the picture across all of them.
What’s yours from here
Everything from the readout is yours: the report, the language for the groups, and the sequence your leaders worked out. Some leadership teams run the sequence themselves. Others bring in partners they already trust, or me. The Diagnostic is complete on its own, and nothing after it is assumed.
Run the same fourteen questions later and compare how people in each function or location describe the work. That can show where conditions improved or slipped, taking account of changes in teams and who responded.
Questions leaders ask
What does it take from my side?
A sponsor on your leadership team. Someone who can send the survey to everyone whose work touches the customer, and tell them it measures the system, not them. About five minutes from each person who answers. Usually two or three people from each group for a conversation. And about an hour with your leadership team for the readout.
What does it cost?
Priced by the size and scope of your operation, and quoted after the first conversation.
Is the survey really anonymous?
Yes. No names, no titles, no manager identifiers. Groups of five or fewer are never reported, on any slice or combination of slices, and small groups are combined. When I ask to talk with people who fit a group’s description, neither of us knows how they answered. I think that’s part of why more than 90% of the people invited answer.
How is this different from what HR already runs?
It measures the system, not the people, so it’s never used to rate a manager or an employee. It compares parts of your operation to each other, not to an industry benchmark.
How do you find the groups, and how do you know they’re right?
The grouping is statistical: k-means clustering on answers to the fourteen questions. It surfaces response patterns, not fixed types of employees or proven causes. AI-assisted analysis helps describe differences worth investigating.
I then examine the flow of work and operating measures, and talk with people who fit each group’s description, across functions and where work changes hands. Your leaders test that picture against what they know in the readout.
What can a leader do with it?
Each business unit leaves with its own reading and a view of how its work connects to the others.
A service leader can see where to investigate before launching a knowledge program, and which teams have practices others may learn from.
An operations leader can examine where a customer promise, decision or measure crosses functions and creates more work downstream. The readout helps both decide where to start, and where to hold off spending until they know more.
Who runs the Diagnostic?
I do, or an experienced practitioner I’ve vetted. Either way, it’s the same fourteen questions, scored the same way. Companies can also license ServiceDNA.
The outcome behind this work is Time to Smile: the moment a customer gets what they wanted, and feels it, earned without needing heroes. The Diagnostic helps your leaders decide where to start.