The Completion Rate Lie

Every L&D lead I talk to says the same thing. They can see who finished the training. They cannot see who understood it.

Over the past several months I have been on calls with L&D managers, sales enablement leads, and training coordinators. Retailers. Restaurant groups. Banks. Disability services. Different budgets, different tools, different problems. Then the same sentence turns up.

I know this gap from the other side. I grew up with speaking anxiety bad enough that I avoided whole situations. When I looked for a way to practise, nothing gave me honest feedback. So I built it. Then I noticed the same gap sitting inside most training programs I came across.

This is the write-up of that October 2026 report, for frontline and sales L&D leaders. It is short on purpose. It puts what I heard next to the best public research I could find. Download the PDF.

The three questions teams cannot answer

1. Did anyone learn this?
2. Where are the gaps?
3. What do we do next?

Read it, then check how many of those you can answer with data.

Almost everyone measures how training felt

This is a synthesis, not a survey. It combines published research from ATD, DeakinCo. with Deloitte Access Economics, and Absorb Software with themes from Elqo’s conversations with L&D, enablement, and training leaders at thirteen organisations. It does not use Elqo product or customer data. Quotes are anonymised to role and industry.

• 93% of organisations use participant satisfaction surveys to assess training quality. ATD, global. Source: td.org, “Beyond the Survey.”
• 42% measure whether behaviour actually changes after training. Same ATD source.
• 51% of Australian organisations say they want better evidence that training benefits business performance. DeakinCo. and Deloitte Access Economics, via technologydecisions.com.au.

Your LMS reports three things. None of them is comprehension.

Completion rate. Time spent. Quiz score. That is the standard dashboard. Ask what each one proves about a new starter facing an angry customer, a compliance question, or a product they have to explain, and the answer is the same. Not much.

What the dashboard says	What it can prove
Completed	Someone reached the last screen.
Time spent	The course was open for a while.
Passed the quiz	Someone recognised a right answer when they saw it.

“they can tick to the end, but when we see the course should take 70 minutes and it’s taken them two, it doesn’t tell a good story when the auditors come knocking”

Head of L&D, financial services

You will often see the claim that people forget most of what they learn within days. It traces back to an 1880s memory experiment using nonsense syllables, and it is regularly misapplied to workplace training. You do not need it. The simpler point holds without it: completion data cannot tell you what stuck.

Teams measure how training felt. They need to know what changed.

ATD’s global figures, from “Beyond the Survey”:

• 93% use participant satisfaction surveys.
• 42% measure actual behaviour change.
• 4% rate themselves extremely proficient at measuring learning impact.

In Australia, among the 68% of businesses that track L&D outcomes, success is measured as satisfaction surveys (56%), productivity change (50%), and financial metrics (33%). Respondents could name more than one method. Source: DeakinCo. and Deloitte Access Economics, via technologydecisions.com.au.

Teams that do reach behaviour and results measurement typically draw on four or more data sources, not a survey alone.

Same problem. Different words.

From discovery conversations with L&D and enablement leaders, anonymised to role and industry.

“people just skip through and we just find they’re not actually learning, they’re just getting to the quiz at the end”

L&D lead, financial services

“if I take 10 minutes to read something and I see people taking two or three minutes, I know they’re just running through it”

L&D manager, building products

“the ability for someone to memorise the right answer versus actually being able to demonstrate certain capability is very, very different”

L&D leader, large multinational

“you can’t actually track where somebody’s up to”

L&D lead, quick-service restaurants

“how do I know if my reps are ready? I don’t.”

Sales enablement lead, technology

“the forgetting curve, people forget... what can we do to support them on the floor so that training is retained”

L&D lead, insurance

Everyone could see activity. Nobody could see understanding. Most already suspected the training was not sticking, and none of them had a way to prove it either way.

One gap. Two job titles.

Frontline L&D

Staff finish the module. Nobody knows if they can handle the real conversation. It shows up as a complaint that goes badly, a compliance answer that is almost right, onboarding handed off and never tracked, or retraining that eats L&D time and still misses.

Sales L&D and enablement

Reps pass product training. Nobody knows if they can explain it to a buyer. It shows up as a pitch that misses what the buyer cares about, inconsistent product messaging across the team, new hires who ramp slower than anyone can explain, and launches that land on slides but not in calls.

• 14% vs about 25%. Average attrition at advanced learning organisations versus laggards. Correlation, not proof that training alone drives it. DeakinCo. and Deloitte Access Economics.
• 65% of sales training and content goes unused by reps. A dated figure, and still the most cited one. SiriusDecisions (now Forrester), 2013.

In the conversations behind this report, the cost was rarely one incident. One sales leader put an underperforming rep at about $45k. A quick-service restaurant group said 75% of new starters leave within 90 days. Neither could trace it back to training, because nobody had the data.

What good looks like

AI is already in L&D. Confidence in the results is not. Absorb Software and Together, in “Enterprise L&D in 2026,” report that 61% of organisations have adopted or are testing AI in their L&D strategy, and 11% feel extremely confident in their approach to building future skills.

A survey asks people how they feel about what they learned. A manager’s gut check covers the few people they happen to watch. Neither scales, and neither is evidence. What changes now is that a person’s understanding can be checked directly, one learner at a time, in their own words.

1. Staff who can explain it, not just recognise it. Picking the right answer from four options is recognition. Saying it clearly to a customer is the job.
2. Managers with individual data, not cohort averages. An average hides the three people who are not ready. A manager needs to know who they are.
3. Gaps found before customers find them. L&D can show the business what is landing and what is not, ahead of the satisfaction scores, not after them.

Three questions every L&D leader should be able to answer

1. If a frontline staff member completed your most important training module last week, could you tell me today what they retained from it?
2. If your training investment is not producing the outcomes you expected, what data would you use to work out why?
3. When your staff meet a real customer scenario, how do you know they are ready for it before they are already in it?

Count the questions you can answer with data, not a guess.

• 3. You are in rare company. Only 4% of organisations rate themselves extremely proficient at measuring learning impact (ATD).
• 1 or 2. You have a gap worth closing.
• 0. You are not alone.

Three things to do before next quarter

If you run frontline L&D

1. Check time against length. Pull completion times for your highest-risk module. If most people finish in a fraction of the expected time, you have your first answer.
2. Ask five people to say it out loud. Pick one module. Ask five staff to explain it in their own words. Listen for what is missing.
3. Ask managers where new starters struggle. Get the two moments that go wrong most often. Build practice around those, not around the whole course.

If you run sales L&D

1. Test one product, one objection. Ask three reps to explain it on a voice note. Score each against what a buyer needs to hear.
2. Find out what reps open. Check which content gets used. Cut what does not, and ask why.
3. Track ramp by conversations. Measure new reps by calls handled well, not by courses completed.

None of this needs new software. Do it by hand first. You will learn what is worth measuring before you buy anything to measure it, and you will have a baseline to compare against later.

The bottom line

Completion tells you training happened. It does not tell you it worked.

The teams that close the comprehension gap first will be able to show the business what is landing and fix what is not, before it costs them a customer or a deal.

Elqo turns the training content you already have into interactive speaking practice. Staff answer out loud and get instant feedback. L&D teams see comprehension data for each learner, not just completions. The designed report is here: The Completion Rate Lie (PDF). If any of this sounds like your team, email me at william@elqo.app. No pitch. If Elqo turns out not to be a fit, I would rather hear that too.

Sources

• ATD, “Beyond the Survey: Why L&D Needs More Than Smile Sheets to Show Impact” (td.org)
• DeakinCo. and Deloitte Access Economics research, reported by Technology Decisions (technologydecisions.com.au)
• Absorb Software and Together, “Enterprise L&D in 2026” (globenewswire.com)
• SiriusDecisions (now Forrester), 2013, via Cirrus Insight (cirrusinsight.com)