AI Readiness Assessment
An AI readiness assessment checks whether your data, handoffs, and ownership rules can support automation before you buy any. OperStack runs seven checks and returns a written score with the two things to fix first, which is often not an AI project at all. Most engagements take One call plus 3 to 5 working days for the written score.
| Entry point | Assessment · Free Free. 30 to 45 minutes plus the written score afterwards. |
|---|---|
| Typical timeline | One call plus 3 to 5 working days for the written score Depends on how much of your data is already clean. |
| Updated | Reviewed by Maksim Shchegolev |
Readiness is about inputs, not ambition
Teams asking whether they are ready for AI usually mean whether the technology is mature enough. That is the wrong question. The technology is ahead of most companies’ data.
The real question is whether the inputs an automation depends on are consistent. If two channels call the same field by different names, a model will happily produce confident nonsense from both, and you will not notice for a month.
The seven checks, and the one that fails most
Capture, identity, ownership, definitions, latency, evidence, decisions.
Ownership fails most often. Not because nobody is assigned, but because at least one routing rule has no fallback branch, so records matching nothing sit with no owner. Nobody sees them because no report counts records that were never assigned.
The reconciliation that surfaces this is described in the inbound lead audit.
Definitions fail second
Marketing counts a lead at form submit. Sales counts it at accepted. Finance counts it at first invoice. All three are defensible and none of them agree, so every meeting starts by reconciling numbers instead of deciding anything.
Fixing this costs one meeting and a written definition, and it is worth more than most automation. The acceptance side of it is in MQL versus SQL.
When we tell you not to buy anything
If your capture layer is inconsistent, fix that first. If your reports do not cause decisions, adding a dashboard makes it worse. If you have fewer than a few dozen inbound records a month, the leak is small enough that a person can watch it and automation is premature.
We would rather say that in week one than sell a setup that measures well and changes nothing. The reporting principle behind it is in inbound lead reporting.
What the engagement covers
- Capture check: does every channel emit the same payload, with consent evidence
- Identity check: can two records for one company be reliably matched
- Ownership check: does every record have a live owner, and every rule a fallback
- Definition check: do marketing, sales, and finance count a lead the same way
- Latency check: where the clock starts, where it pauses, and who watches it
- Evidence check: could you reconstruct why a specific lead was handled the way it was
- Decision check: does any current report actually cause a decision
What you keep
- A written score per check, with the evidence we used to reach it
- The two highest-cost gaps, ranked by what they cost you rather than by effort
- A short recommendation, including when the recommendation is to do nothing yet
What this does not cover
- A vendor shortlist. We are not a reseller and we do not rank tools for you
- A maturity model with five levels and a logo. The output is two problems and a number
- Access to your production systems. Read-only exports and a call are enough
If one of these is the actual problem, say so on the audit call and we will point you somewhere better rather than sell you the wrong module.
Start with the written assessment
Free and in writing: send the site and how inbound works today, get the gap we see, the module that fixes it and a fixed price within two working days. No call at this stage.
Frequently asked questions
- What does an AI readiness assessment actually measure?
- Whether the inputs an automation would depend on are consistent enough to automate. Payload shape, identity matching, ownership, shared definitions, timing, and evidence. Model choice is not part of it, because model choice is rarely the constraint.
- Is it really free?
- Yes, and there is no obligation afterwards. Scoping without seeing data is guesswork, so the assessment is as much for us as for you. Roughly speaking, if we cannot see the leak we cannot quote the fix.
- What if the answer is that we are not ready?
- Then that is the answer and we say it. Usually the blocker is a capture layer emitting different field names per channel, which is a week of work and not an AI project.
- What do you need from us?
- A read-only export of recent inbound records, a look at your form and routing configuration, and forty-five minutes with someone who knows how leads actually get worked. No production access.
- How is this different from a sales call?
- The output is a written list you keep whether or not you hire us. If the fix is inside your existing tooling, we will tell you which setting to change rather than quote for it.