Hello Data-DrivenRecruiters! Have you lost weight? Crossfit? well good for you!
Today we’re talking ATS selection.
This is something I’ve written about a lot over the years because it is just such an ever present risk for changing recruitment functions.
The ATS market used to be one of the most underserved areas of business software.
That is absolutely no longer the case.
In the last five years it has completely exploded.
We now have:
- Big legacy players with massive market share Newer platforms shipping features at an insane pace Smaller tools solving very specific problems extremely well
We have never been more software-enabled as a function.
Which sounds great.
But it also means choosing the wrong ATS is now easier than ever.
So here are five tips to help you not buy a shit one.
1) Process first.
If you haven’t written out your recruitment process properly, you are not ABLE to choose an ATS in any meaningful way. You are guessing.
Listing stages does not count.
I don’t mean:
Screen → HM → Panel → Offer
I mean the actual process.
A full flowchart for every step recruiters take.
Every step candidates experience.
Who does what.
Who waits for what.
Where delays happen.
How feedback is captured.
What format information is in.
Emails, calls, forms, documents, NDAs, info packs, references, onboarding handovers.
All of it.
They tell you exactly where your system is getting in the way and what would actually improve your outcomes.
Until you’ve mapped this visually using a white boarding tool, you do not know what you need from an ATS.
Once you do, the requirements become obvious.
And suddenly you can say:
“I need X and Y. I absolutely do not need A and B.”
That is when vendor conversations start being useful.
2) Scale is mostly not a consideration.
I hear this constantly:
“We’re getting bigger, so we need a more enterprise ATS.”
In most cases, that statement does not survive five minutes of scrutiny.
Almost every modern ATS can handle more roles, more candidates, and more users.
The system is not going to fall over because you hired more people.
What people are really describing is this:
“We want to increase hiring output without increasing recruiter headcount.”
That is an automation and workflow problem.
If your ATS does not support automation properly, it was never a good system to begin with. That would have been true at one hire or one thousand hires.
3) Proper Governance
This one hurts because I’ve seen it so many times.
TA teams do the work.
They shortlist vendors.
They run trials.
They gather feedback.
And then at the eleventh hour, the CPO goes "good news, Workday said they'd throw in the ATS for 8 grand!
Then you enter a pit of despair.
Especially the candidates filling out 17 page application webforms that were built circa 2004, just to get a part time janitor role.
Before you even speak to vendors, write down governance.
Who is accountable for the decision.
Who is responsible for running selection and implementation.
Who needs to be consulted.
Who just needs to be informed.
As a general rule:
- TA leadership should be accountable TA Ops and implementation owners should be responsible HR, Finance, IT, Legal should be consulted The wider business should be informed
If this is not explicit, you can still end up with the wrong tool even if you did everything else right.
4) Interrogate the AI claims early and aggressively
Everything has AI now.
Fine.
Ask them to show you exactly what it does.
Live.
With your use case.
Ask what triggers it.
Ask what data it uses.
Ask how it fails.
Ask what control recruiters have over it.
Ask how much time it actually saves per week.
If a vendor cant clearly explain where the value appears in your process, there is no value.
A tool doing something badly at speed is just faster chaos....that you pay for.
If you cant clearly map the feature back to your process, stop the demo and go back to tip number one.
5) Do not overcomplicate implementation
You do not need every feature turned on from day one.
That is how you burn out recruiters and yourself
Figure out your MVP.
Implement that properly.
Get adoption.
Then layer in complexity over time.
Treat it like a product rollout, not a big-bang transformation.
This approach is better for outcomes, better for adoption, and significantly better for the old mental health.
Luke's Final Thought
If you want to avoid buying a shit ATS:
- Understand your real process Ignore scare stories about scale Lock in governance early Push hard on AI claims Implement in small, deliberate steps
Well... that's issue fifty seven! If you have any questions about it, or any feedback on this issue of The Data Driven Recruiter, grab me on LinkedIn for a chat.
I'll see you next week!