Employee Experience Consulting: What You're Buying and When to Skip It

Three proposals sit on your desk and there’s no way to line them up.

One is a survey program ending in a diagnostic report. One is a platform rollout with a per-seat license attached. One is a slide deck about your operating model, priced above both. Each answers a different question, none the one you asked on the first call, which was roughly: people are having a worse time here than they should be, and I need to know what to do about it.

Employee experience consulting is one search term covering at least four products, with different deliverables, different demands on your team, and very different things left in your hands. Here’s what each one is, when you shouldn’t buy any of them, and the questions that separate a firm worth signing from one that will hand you a PDF.

What employee experience consulting delivers

The work covers what it’s like to work at your company, from the day people join to the day they leave. Firms sell four things against that. Measurement. Design. Software. Sequencing.

The promise is easy to write, so look past it to the artifact: what lands on your desk, and what it lets you do the Tuesday after. Most of the money still goes to measurement, because a survey produces something visible fast and a report is easy to price. A report is also the hardest artifact to convert, since a finding sits where you left it until somebody with budget and authority moves.

One number sits under all four: employee lifetime value. How well people perform, and how long they stay. Culturro has worked on those two things since 2017, and they make a fair test for anything here. Work that moves neither performance nor tenure is protecting you from a downside, fine as long as somebody says so out loud. The chain from an HR lever to a number finance reports is in how HR contributes to business growth.

What changed once AI arrived in the category

Three things moved, and one didn’t.

Free text at scale is the real change. Comment fields that used to get sampled by an analyst on a deadline now get read in full, clustered and summarized in an afternoon. It works. It also puts your employees’ unfiltered writing about their managers through a model somebody selected, which makes an HR purchase a CISO question: where it runs, what’s retained, what trains on what.

The deliverable can now be a system, and that changes the ownership conversation. Owning a deck is trivial. Owning a system means code, prompts, the evaluation set that proves it behaves, runbooks, and a pipeline that keeps running long after the invoice clears.

The seller mix changed too. One search term now returns people firms that subcontract the engineering and engineering firms that added HR last year. Ask which one you’re on the call with, and where the other half comes from.

What didn’t move is who decides. AI can read, cluster, rank, summarize and draft. A person makes every call about a person, and the build should leave no path around that. Regulation has caught up. Under the EU AI Act, AI used for recruitment, promotion, termination and task allocation is high-risk, and Article 22 of the GDPR gives people a right not to be subject to decisions based solely on automated processing where the effect is significant. Press hard wherever a proposal goes vague on this. The listening tools sit on a map too: 10 engagement and listening use cases, with who decides on each.

The four engagement types, compared

Engagement	What lands on your desk	Who it suits	What it takes from your team	What you own at the end
Diagnostic and survey	An instrument, results cut by segment, a findings report	You feel a problem and can’t name it, or you need evidence a skeptical exec team accepts	Comms, scheduling, chasing response rates, interview slots. Light hours for HR, visible to everyone	The data, the instrument if you negotiated for it, a report. The changes stay yours to make
Program design	Designed processes: onboarding, manager rituals, career frameworks, listening cadence, plus rollout materials	You know the problem, you’re short on design capacity, and your managers will run what you hand them	Workshops, review cycles, and a lot of internal selling to get it accepted	The design and the materials. Adoption sits with your line managers from day one
Platform implementation	A configured product, integrations, dashboards, admin training	You’ve bought or are about to buy software and want it configured and adopted properly	Data cleanup, HRIS access, an IT and security review, a named admin who stays after go-live	A configured tenant. The platform stays the vendor’s, and your process bends toward their data model
Strategy, then build	A sequenced roadmap with a business case per item, then working systems built against it	Several competing ideas, board pressure, and no order you could defend	A handful of working sessions of about two hours, one data pull, a named owner per number	The roadmap, and whatever gets built after it, running in your environment

Each fails in its own way, and the failure mode is what to buy on.

Diagnostic and survey work

The report arrives, everybody agrees with it, and nothing happens, because the recommendations sit at a level no single person can act on. Ask before signing: who receives this, what decision does it feed, and what’s in next quarter’s plan that changes because of it. If you ran a survey in the last year and haven’t finished acting on it, another one spends credibility with people who already told you.

Program design

The design assumes a manager population you don’t have. Career frameworks need managers who can hold a career conversation. A listening cadence needs managers who act on what they hear. Strong design plus unready managers gets you a beautiful document and a quiet rollout. Ask what the design assumes about your managers, and what happens if that’s wrong.

Platform implementation

Cost of ownership arrives late. License renewal, the admin role nobody budgeted for, the moment three years in when your process bends to the roadmap of a company you don’t control. It’s the right buy when the product fits and you want somebody who’s done that configuration twenty times. Price the whole arc, including the year it takes to leave.

Strategy, then build

You wait longer to see anything, and a month spent sequencing feels expensive when the board wants progress this quarter. It earns its place when building the wrong thing first would be costly, which in HR is close to the default, since the bill for a system people distrust arrives as behavior nobody logs. It’s the wrong shape when you know what to build and who owns it. That’s a build, and you should pay for a build.

That last one is what we sell, so read it knowing that.

When you don’t need a consultant

A firm worth hiring turns work down. Here’s where we’d tell you to keep the money.

1. You can already name the problem, the fix, and the owner. What’s missing is permission. Paying an outside firm to repeat what you’ve said does work, and it’s expensive. Cheaper: get the function head who owns the number to say it in the room, standing next to you.
2. It’s one team, or one manager. Named-person problems don’t respond to programs. That’s an HRBP conversation this week, and a consultant takes three months to say the same thing with a chart.
3. The real constraint is pay, headcount or workload. No experience program survives an understaffed team or a band below market. Fix the input. Anything else reads to employees as a distraction, and they’ll be reading it correctly.
4. The exec team hasn’t agreed there’s a problem yet. A report commissioned into an unsettled argument becomes ammunition for whoever likes its conclusions. Settle that first.
5. There are open commitments from the last piece of work. Close them. Restarting a listening cycle with promises outstanding costs trust that comes back slowly.
6. The plan already exists and what’s missing is capacity. Roadmap set, decisions made, owner named: hire contract hands or a permanent person. Both cost less per hour than advice. If it’s a system you want, with clean data behind it, skip the strategy step and go get a build quote. We say this on first calls.
7. You’re small enough to ask everyone yourself. Below a couple of hundred people, a CHRO who spends two weeks in real conversations gets better material than an anonymous survey produces. What a consultant sells is reach across a population you can’t personally cover.

None of these mean the work isn’t worth doing. The money just does more somewhere else.

What to ask any firm before you sign

Four areas, all askable on a first call.

Who actually does the work

Names of the people who’ll be on it, hours per week, what else they’re carrying. Whether the person in the pitch is on delivery or just in the room. Which parts get subcontracted, and to whom. A firm that can’t say who’s writing the code is selling a team it hasn’t assembled.

Where the data sits

Where the system runs and in whose account. What’s retained and for how long. What trains on what, with the model provider’s no-training and retention terms in writing. Who at the firm can see employee-level records, and what happens to that access on the last day. Ask what standard they work to: some map their process to the NIST AI Risk Management Framework, some to ISO/IEC 42001. A firm that answers all five from memory has done this before.

What you keep

For a report: the raw data and the instrument. For a design: source files and the right to change them without asking. For a system: code, prompts, evaluation sets, documentation and runbooks, in the contract, with no dependency on the firm’s platform to keep running. The clean version is, what breaks if we never speak again.

What happens when it fails

Where the go/no-go points are, and what a written no looks like. Ask for a time they recommended against building something, and listen for whether that client stayed. Ask what the support period covers and what gets billed afterwards. Then ask what they’d have you measure to conclude the engagement didn’t work. That one produces the most informative five seconds of the call.

Where we sit in those four types

We’re a people consultancy that learned to build software. Culturro started in 2017 on HR strategy, policy, engagement, training and leadership development, across 30+ organizations. HR AI is the sixth practice and the current one. We run software too. Agnya is our employee experience platform. Merlin, the coach inside Risely, our leadership development product, covers 83 workplace skills in 40 languages and has worked with 5,000+ people from 40+ companies.

Strategy first, on one fixed fee: about a month to answer five questions for your function, ending in a written go or no-go on whether anything is worth building. The roadmap and business case are yours either way. Each use case after that is priced on its own: a fixed fee for a working proof on test data, quoted from the roadmap, then a fixed fee for the agreed build scope, quoted once you’ve seen that proof run. A support period after handover is included, scoped with the build. Hosting and model usage get billed to you by your own providers. Ongoing support is a month-to-month retainer, offered after the strategy step, and a delivered system runs without it. Code, prompts, evaluation sets, documentation and runbooks transfer to you at handover.

We don’t build systems that make autonomous decisions about people, and what AI won’t decide in your function gets written down in the strategy step. “Not yet” is a real output here, with the one thing that would change the answer and a date to look again. Put the four question sets above to us too.

Start with the question you’d pay someone to answer

Before you compare those proposals, write down the number you’d want moved twelve months from now and the person outside HR who reports it. Some of them stop making sense the moment that’s on paper.

Then take the AI idea already pitched to you and score it against the 10X HR-AI Framework. It runs to eight questions and ends on a tier, one of which is a straight no for now. Holding a queue? Three scored end to end shows the order that falls out.

For the whole list sequenced with a business case attached, that’s the strategy month: the steps, the commercials and what we ask of your team, written out in full.

Frequently asked questions

What does employee experience consulting cost?

There’s no standard price, and its shape tells you as much as the number. Day rates put the risk of a long job on you. Per-seat licenses grow with headcount whether or not anyone uses the thing. A fixed fee for a defined scope puts the overrun on the firm. Ask what would make the number change before you ask what it is. Ours is a fixed fee at every step: the strategy month, then each use case, with the build quoted after the proof.

Can we do this in house?

Yes, in many cases. Outside firms bring three things internal teams can’t: pattern exposure across companies, time not committed to running the function, and permission to say something unpopular. If your team has all three, do it yourselves and spend the budget on the change.

Should this sit with HR or IT?

HR owns the outcome and the call on what stays human. IT and security own where the data sits and what it can touch. Engagements stall when HR runs a procurement alone and meets the security review in month three. Bring your CISO into the first conversation and those questions get answered while they’re cheap.

We already bought an employee experience platform. Now what?

Pull the adoption numbers and support tickets before buying anything else. A platform nobody opens is a configuration problem, a manager problem, or a trust problem. Only one of the three is solved by more software.