Enablement's ROI Problem Isn't a Measurement Problem. It's a Naming Problem.
You probably don't have an attribution problem. You have reps behaving differently after a program and nobody on the enablement team who wrote down which specific competency changed.
Every enablement leader I've sat across from has said some version of the same sentence: "we know it's working, we just can't prove it — attribution is really hard." I've stopped believing the second half of that. Attribution isn't hard. What's actually missing is a noun: the specific competency the programme was meant to close. No metric on earth can attach itself to a programme that never named one.
The tell
You can spot a naming problem before you look at a single number. Read the name of the last three enablement programmes you ran. If they're named after a delivery format or a calendar event — "Q3 Kickoff Deep Dive," "Negotiation Bootcamp," "Discovery Skills Refresh" — you've named the wrapper, not the skill. Ask the manager of a team that went through one, six weeks later, what specifically changed, and you'll get "reps seem sharper" or "the team feels more confident going into negotiations." That's a vibe. It is not a competency, and vibes don't show up as line items.
I've sat in the QBR where this plays out. Head of enablement stands up, says the Q2 negotiation programme was a hit, cites a satisfaction score of 4.6 out of 5. The CRO asks what changed in actual deals. Silence, then something about reps reporting more confidence going into pricing conversations. That's the exact moment the enablement budget gets flagged for review next cycle — not because the training didn't work, but because nobody in the room, including the person who ran it, can say what it was supposed to change specifically enough to check.
Why "attribution is hard" is the wrong diagnosis
Here's the test. Take a competency that is actually named and levelled — say, "Economic Objection Reframing," four levels, with Level 3 defined as something observable: the rep reframes a stated price objection around cost-of-inaction without getting defensive, inside the same call. You can grade a sample of reps against that definition before a programme runs. You can grade the same reps again after. You can isolate the ones who moved up a level and check, specifically, what happened to their win rate on deals where that exact objection type came up — not deals in general, the ones where the skill was actually in play. None of that is hard measurement. It's ordinary measurement, applied to a target that was defined precisely enough to be measured. Most programmes skip the naming step because it's slower and considerably less fun than booking the workshop, and then call the resulting silence an "attribution problem," which sounds like a sophisticated limitation rather than what it actually is: homework that didn't get done.
Before and after naming it
| What it's called now | What's actually being asked to change | What you can measure once it's named |
|---|---|---|
| "Negotiation Bootcamp" | Unclear — could be discount defence, could be multi-stakeholder trade concessions, depends who's teaching it | Nothing durable; every rep interprets the ask differently |
| "Discount Defence, Level 2 → 3" | Hold margin against a named competitor's price without offering a first counter under 5% | Win rate and average discount on deals with a documented competitor mention |
| "Discovery Skills Refresh" | Unclear — a two-hour session covering open questions in general | Nothing durable; no observable behaviour was defined |
| "Economic Impact Quantification, Level 1 → 3" | Rep surfaces a quantified cost of inaction before the proposal stage | Share of closed-won deals with a documented, quantified business case pre-proposal, and cycle time on those deals |
Notice what the named version buys you beyond the board slide. Once you can point to a specific rep moving from Level 1 to Level 3 on "Economic Impact Quantification," and to what happened to their cycle time as a result, you also have something real to say in that person's own development conversation — not "you were great in the workshop," but "this specific thing you now do differently is worth this much, and here's the next level up." The aggregate ROI number and the individual coaching conversation turn out to be the same piece of work, done once, instead of two separate exercises where one is guesswork and the other is flattery.
What naming actually requires
- Name the single skill, not the topic area or the event. Use a levelled definition — something like the Sales Certification Rubric (Novice-to-Expert Levels) — so "better at this" has an actual boundary rather than a feeling.
- Establish a real pre-state. Grade a sample against the rubric before you run anything, or have reps run a Skill Gap Self-Assessment Quiz first, so you're comparing against a measurement, not a memory of how the team seemed a quarter ago.
- Define the observable behaviour at the target level precisely enough that two different graders, listening to the same call cold, would score it the same way. If your definition can't survive two graders disagreeing, it isn't a definition yet.
- Decide the downstream metric in writing before the programme runs — the specific metric that specific behaviour plausibly touches, over a stated window. Not "revenue," which touches everything and therefore proves nothing. The metric that this exact skill, and only this skill, would plausibly move.
- Re-grade on the sales cycle's timeline, not the training calendar's. Most programmes measure at the 30-day mark because that's when the calendar reminder fires, not because that's when a competency actually shows up in live outcomes. If your cycle is 90 days, measuring at 30 tells you almost nothing and gets reported as a null result that wasn't actually null — it was just early.
None of this is a measurement innovation. It's paperwork done before the workshop instead of during the retro, when the numbers didn't move and everyone's trying to explain why in terms vague enough that nobody's specifically wrong. Name the skill first, and the ROI conversation stops being a defence you have to mount after the fact. It becomes a fact you already had on file.