eCommerce Audit Template
Most audits do not fail at the checking stage. They fail at the handover, when forty unranked observations arrive with no severity, no effort estimate and no owner, and the person receiving them acts on none of them. This workbook is the part that comes after the checking.
FORMAT
XLSX
CONTAINS
4 tabs
LAST UPDATED
September 2026
LICENCE
Free, unrestricted
An audit that hands over forty tasks hands over none.
- Why this one
Our checklists tell you what to check. Nothing told you what to do with the answers.
This is the missing half, and it is the half that decides whether an audit changes anything.
The difference between an audit and a list of complaints is a severity column and an owner.
There are already four free checklists on this site covering what to look at. What none of them do is hold the output. So an audit produces a document listing forty problems, the person receiving it reads the first six, and the work that follows is whichever finding was described most dramatically rather than whichever one is costing the most.
The fix is unglamorous: score every finding on severity, confidence and effort, let the sheet rank them, and hand over ten rather than forty. The priority column is deliberately crude arithmetic. Its job is to stop the loudest finding being treated as the most important one.
- One row per finding, with the evidence beside it
- Severity, confidence and effort on scales with a test for each level
- Priority calculated, not argued about
- Impact sized only where you can show the arithmetic
- A coverage tab that records what you did not audit
Two counters exist purely to catch an audit before it is presented: findings with no evidence, and findings with no owner.
- Where this sits
Four things on this site have the word audit in them
They do different jobs. This is which one you want.
01
THIS WORKBOOK
The output layer. Use it while running any of the checklists, or while auditing in your own way entirely.
02
THE CHECKLISTS
Nine of them, forty-two checks each, covering SEO, CRO, product pages, paid media, technical and more. Free, and the input to this workbook.
03
THE ASSESSMENT
Twelve commercial checks with published thresholds, aimed at the question of where the money is leaking rather than at a complete inventory.
04
THE SERVICE
Five working days, the specialists who run each discipline, a ranked constraint list and a ninety-day sequence.
- What is inside
Four tabs
Small on purpose. A findings workbook with thirty columns does not get filled in.
Read me
Six summary counters and the rule the workbook exists to enforce: rank before you recommend.
- Findings logged, and how many are severity 4 or 5
- Quick wins: high severity at low effort
- Total sized impact, labelled as deliberately incomplete
- Findings with no evidence, and findings with no owner
Findings
Fourteen columns, fifty-nine rows, and a priority score on every one of them that calculates from the scores you enter.
- Reference, discipline, where found, the finding itself
- Evidence, in its own column, because a finding without it is an opinion
- Severity, confidence, effort, and calculated priority
- Annual impact, owner, target date, status and notes
Scoring
Three scales with a plain test for each level, so two people auditing different disciplines produce comparable numbers.
- Severity 1 to 5, from untidy to measurably costing money
- Confidence 1 to 5, from hunch to reproducible
- Effort 1 to 5, from hours to a programme
- How to size impact, and when not to
Coverage
Six disciplines, what a complete pass looks like in each, and a column to record whether you actually did it.
- Acquisition, conversion, retention, merchandising
- Technical, and operations and margin
- Which of our checklists covers each one
- A record of the gaps, so silence is not read as a pass
Nothing is locked or protected. If the priority formula does not suit how you work, change it.
- The scoring
Three scales, each with a test
Scores invented as you go are not comparable with each other, which is why the tests matter more than the numbers.
Score
Severity means
Confidence means
Effort means
5
Costing money today, measurably
Measured and reproducible — anyone could open the same report
A programme: replatform, reorganisation, supplier change
4
Costing money, but not yet measured
Measured once, plausibly stable
A project: more than one team, or a dependency you do not control
3
Limiting growth rather than losing revenue
Strong inference — related numbers move as expected
Weeks: a developer, a staging environment, testing
2
A risk rather than a loss, so far
Professional judgement you cannot show here
Days: one person, a template change, a release
1
Untidy. Fix it when somebody is in there anyway
A hunch. It should never outrank anything
Hours: one person, no approval, no deployment
Priority is (severity x confidence) divided by effort. Crude, on purpose, and better than whichever finding was described most loudly.
- How to use it
Running an audit that survives the handover
The scoring takes minutes. It is the thing that makes the rest of the days worth anything.
01
Agree the scales before you log anything
Read the Scoring tab first. If two people are auditing different disciplines and scoring severity differently, the ranked list at the end is meaningless and nobody will be able to say why it feels wrong.
02
Log findings as you go, with the evidence beside them
Not afterwards from memory. The evidence column is the one that turns a finding into something a client or a board cannot wave away, and it is the first thing missing when findings are reconstructed later.
03
Score honestly, including confidence
Confidence is the column people are tempted to inflate. A finding you believe but cannot show scores a 2, and the formula will correctly rank it below something you can prove. That is the point.
04
Size impact only where you can show the arithmetic
Put the working in the Notes column so somebody can disagree with it. The summary total says out loud that it is incomplete, which is more defensible than a total built partly from guesses.
05
Sort by priority and hand over ten
Not forty. The other thirty stay in the workbook and get picked up in the next pass. An audit that hands over everything transfers the prioritisation problem to somebody who has less information than you do.
06
Fill the Coverage tab even for what you skipped
An audit silent on retention reads as an audit that found nothing wrong with retention. Recording the gap is more useful than the appearance of completeness.
Re-run the top ten in ninety days rather than starting a fresh audit. Progress against a known list beats a new list every quarter.
- Getting it wrong
Six ways audits fail after the checking
Every one of these is a handover problem rather than an analysis problem.
Handing over an unranked list
Forty findings with no order transfers the hardest decision to the person least equipped to make it. They act on none of it and conclude the audit was not worth paying for.
Letting the loudest finding win
The finding described most dramatically is rarely the one costing the most. A severity column exists to make that visible rather than arguable.
Inventing impact figures
A number you cannot defend discredits the findings around it. Leave it blank, say the total is incomplete, and keep the credibility of the ones you can show.
No owner on anything
A finding with no name against it has no failure mode, which is why it is still there at the next audit.
Not recording what you skipped
Silence on a discipline is read as a clean bill of health for it. Operations and margin is the row skipped most often, and the one most likely to hold the answer.
Adding up the impacts
Two findings on the same journey rarely both deliver. Fixing the first removes part of the second opportunity, so the total overstates what you will actually get.
If you recognise the first one, that alone is worth the ten minutes this workbook costs to set up.
DOWNLOAD
eCommerce Audit Template — XLSX
A findings log with fourteen columns and a priority score that calculates itself, three scoring scales with a test for each level, a coverage tab across six disciplines, and six summary counters including the two that catch an audit before it embarrasses you.
FORMAT
XLSX spreadsheet
TABS
Read me, Findings, Scoring, Coverage
FORMULAS
65, all working
LICENCE
Free, commercial use
No email address, no sign-up and no watermark. Three worked example findings sit at the top of the Findings tab, each labelled, and should be deleted before you start.
- Questions
Audit template FAQs
What people ask before downloading it.
Is this a checklist?
No, and deliberately so. The checklists tell you what to look at and they are free. This is where the answers go, so that what you found becomes a ranked list rather than a document nobody acts on.
How is the priority score calculated?
Severity multiplied by confidence, divided by effort. It is crude on purpose. A more sophisticated model would invite arguments about the model rather than about the findings, and would not change the top ten much.
Why is confidence a separate score?
Because it is the honest brake on severity. A finding you are certain about but that only limits growth should usually rank above one you suspect is costing money but cannot show. Without a confidence column, the person with the strongest opinion sets the order.
Should I always size the annual impact?
No. Size it where you can tie it to a measured figure and show the working in the Notes column. Everywhere else, leave it blank. The summary total states that it is incomplete, which is far more defensible than a total built partly from guesses.
Can I use this for a client audit?
Yes, commercially and without attribution. Many people will want to hide the Scoring tab before sending it on; we would leave it in, because showing the client how findings were ranked is usually what makes them trust the ranking.
How many findings should I hand over?
Around ten. The rest stay in the workbook for the next pass. Handing over forty transfers the prioritisation problem to somebody with less information than you have, and the predictable result is that nothing happens.
What is the Coverage tab for?
Recording what you did not audit. An audit that says nothing about retention is read as an audit that found nothing wrong with retention, and the gap is invisible to whoever receives it unless you write it down.
Does it work in Google Sheets?
Yes. Upload the XLSX to Drive and open it with Sheets. The formulas are IF, OR, SUM, COUNTA and COUNTIFS, which behave identically in both.
- Keep going
Related reading
CHECKLISTS
Nine checklists, forty-two checks each. The input to this workbook, and free.
FRAMEWORK
Twelve commercial checks with published thresholds, for finding the one binding constraint rather than a complete inventory.
TEMPLATE
Where the operations and margin row of the Coverage tab gets its numbers.
Rather not run it yourself?
The free audit covers all six disciplines with the specialists who run each one, and arrives as a ranked constraint list with a ninety-day sequence rather than a list of forty things.