Why recruiters only use a tenth of their CRM
Most of what you pay for was never switched on, never appeared on the screen a consultant lives in, or was taught once on a day nobody could absorb anything. All three are fixable without changing system.
Ask AI about this
Ask an agency owner how much of their CRM gets used and you will usually hear a rueful guess of about ten per cent. Ask the consultants and they will describe a working day that runs on a list view, a record page, a notes box and the email window, with everything else reached occasionally when somebody remembers it exists.
This is treated as an embarrassing secret and it should not be. It happens at nearly every agency, on nearly every system, and the reasons are mechanical rather than cultural. Once you can see the places where capability gets lost between the price list and the screen, you can go and get a decent part of it back.
Why does so little of a CRM get used?
Capability gets lost at four separate points, and by the time it reaches a consultant's day very little of what you bought is left. It has to exist in your plan, then be switched on and configured for how you work, then be visible from the screen where the work actually happens, then be part of a routine somebody enforces. Anything that fails at any one of those four is paid for and invisible.
The largest losses are usually at the middle two. Plenty of function is included in the licence and sitting behind a settings page nobody has opened since go-live. Plenty more is switched on but reached by a route no consultant would ever discover, which is the same as being off.
The fourth point is the one owners underrate. A feature that produces value only when everybody uses it, such as structured pipeline stages or logged call outcomes, stays dead until someone in charge insists, because a consultant who logs their calls properly while nobody else does gets nothing for it.
Training happens once, on the worst possible day
Almost all CRM training lands in the fortnight around go-live, which is exactly when nobody can absorb it. Consultants are trying to find their candidates in an unfamiliar system while still hitting their numbers, so they learn the five things that get them through the day and stop there. Those five things become the system forever.
What is missing is the second training, three months later, when people have a working mental model and a list of things that annoy them. That session teaches a completely different set of things, because now the questions are real. It costs a couple of hours and it is the highest-return training an agency ever buys, and it is skipped almost universally because go-live training felt like it was the training.
A related habit is worth building. When a new person joins, they get taught by whoever is free, which means they inherit that person's tenth rather than the system. Written notes on how your agency does the ten most common tasks, kept current by whoever knows, cost very little and stop each generation of hires from inheriting a narrower version of the tool.
Nobody owns the configuration after go-live
At go-live, configuration has an owner, a deadline and a project. The week after, it has nobody. Stages, templates, fields, reports and automations sit frozen in the shape your agency had on the day you switched, while the agency itself keeps changing.
The fix is dull and it works. Name one person as the owner of the system, give them an hour a fortnight, and have them do three things: read what the vendor has shipped, sit with a consultant and watch them work, and change one thing. Not a project, one thing. A template that is out of date, a field nobody fills in, a stage that no longer matches how you sell.
Vendors ship constantly now and the shipping is invisible unless somebody reads it. We publish a dated changelog every week for exactly this reason, and an hour a fortnight spent reading your own vendor's is the cheapest capability you will ever add, because you have already paid for all of it.
What actually predicts whether a feature gets used
A feature gets used when it is reachable from the screen where the work already is, and it dies when it requires a detour. A consultant on a candidate record who can send the CV, book the call, text, start the sequence and log the outcome without leaving that record will do all five. Move any one of them to a separate area of the system and its usage falls off immediately, no matter how good it is.
This is why comparing feature lists between vendors is such a weak way to buy. Two systems can both have a texting feature and only one of them will be used, because in the other it lives somewhere that would require a consultant to decide to go there.
The same logic applies to your own configuration. When you are deciding where to put something, put it where the hand already is. A slightly worse tool on the record beats a better one behind a menu every single time.
| What you see | Where the capability was lost | What to do about it |
|---|---|---|
| Consultants export to a spreadsheet | Reporting was never configured | Build the three reports they rebuild by hand |
| Everyone uses their own email templates | Shared templates went stale | Owner refreshes the top five, delete the rest |
| Calls logged as a note, or not at all | Logging takes more steps than it saves | Cut it to one click and then insist on it |
| A feature you know you pay for is unused | Switched on, but off the main screens | Put it on the record page, or drop the module |
| New joiners work differently from everyone | Training passed hand to hand | Ten written tasks, one owner keeping them current |
| Nobody knows what shipped this year | Nobody reads the changelog | One hour a fortnight, one change made |
How to find out which tenth your agency uses
Sit behind one good consultant for two hours and write down every screen they open and every place they type something that already exists elsewhere. That single exercise tells you more than any usage report, because it shows you the workarounds, and workarounds are where unused capability is hiding.
Then ask your vendor for whatever usage data they hold. Most can tell you which features your account touches and which it never has. Put that list next to your plan and you will have a short column of things you pay for and do not use, which is an agenda rather than an accusation.
Finish by asking your consultants a narrow question rather than a broad one. What did you do more than five times yesterday that felt like typing something twice. Broad questions about the CRM get you complaints about the CRM, and that specific one gets you a list you can act on.
What good looks like a year in
An agency using its system properly has a named owner, a fortnightly hour, a second training session behind it, and a short written record of how the ten common tasks are done here. The tenth turns into something like a half, the software stops being a filing obligation, and the argument about whether to switch system quietly goes away in about half the cases I see.
My own view, after a lot of these conversations, is that most agencies switch CRM one system too early. The pain is real and the cause is usually configuration that stopped being tended and training that happened once at the worst moment, both of which travel with you to the next system. Spend a month proving the problem is the software before you spend a year proving it was not.
Anshika leads Customer Success at Recruitly.
