Run these five checks first. Flow coverage, exclusion logic, list health, segmentation quality, and attribution settings. They take about 30 minutes and they will tell you most of what an agency's paid audit would, which means you walk into the sales call able to tell a real diagnosis from a rehearsed one.
This is not about replacing an agency. It is about being able to evaluate one.
Why bother auditing before the call
Agency audits are a sales tool. That does not make them dishonest, but it does mean they are built to surface findings that justify the engagement being proposed. If you have no independent read on your own account, you cannot tell whether the problems presented are the biggest ones or simply the ones that fit the service being sold.
The asymmetry disappears the moment you have looked yourself.
Check 1: Flow coverage and revenue concentration
Sort your flows by revenue. Write down revenue per recipient and conversion rate for each.
Then check whether the core set exists and is live: welcome, abandoned checkout, browse abandonment, post-purchase, and a winback or sunset path. Klaviyo publishes its own flow audit checklist covering this ground.
What you are looking for is concentration. If one flow produces the overwhelming majority of flow revenue, the account is not a program, it is a single working automation surrounded by decoration.
For context on whether flows are pulling their weight at all, Klaviyo's benchmark data across more than 183,000 brands shows flows producing roughly 41% of email revenue from just 5.3% of sends, with revenue per recipient close to 18 times campaign levels. If your flow share of email revenue is far below that, the gap is your opportunity estimate.
While you are in each flow, check the entry counts over the last 90 days. A flow with revenue in its history and near-zero recent entries has silently stopped triggering, usually after an integration or platform change, and nobody noticed because the dashboard still shows lifetime numbers.
Check 2: Exclusions, which is where most accounts break
Open each flow and read the filters rather than the emails. Three questions:
- Do purchasers exit the welcome flow when they buy, or do they keep receiving a first-purchase discount they no longer need?
- Are active subscribers excluded from replenishment and winback flows?
- Do campaigns and flows overlap, so someone can receive a promotional campaign and an abandoned checkout reminder for the same product on the same day?
Exclusion logic is invisible in a revenue screenshot and it is where the majority of real problems live. An agency that walks you through exclusions unprompted is a good sign.
Check 3: List health, with the thresholds that now have teeth
Build a segment of profiles with no click, site visit, or order in 180 days and compare it to your total mailable list. If that group is a large share of your list and you are still mailing them, deliverability is being spent to reach people who stopped paying attention.
Use clicks rather than opens for this. Apple Mail Privacy Protection registers false opens on delivery and inflates open rates by roughly 15 to 35%, which makes an open-based engagement segment quietly meaningless on any list with Apple users.
Then check your spam complaint rate against the numbers that are now enforced rather than advisory. Klaviyo's guidance is to stay well under 0.1%, and since 2024 Gmail and Yahoo reject bulk senders outright at 0.3%. While you are at it, confirm SPF, DKIM, and DMARC all pass on a real send. If you are near the line, no amount of better copy fixes the numbers, because the problem is who you are sending to.
Check 4: Segmentation quality
Look at how your segments are defined. Good segmentation is built on purchase behavior, engagement recency, and category affinity. Weak segmentation is built on arbitrary time windows that do not reflect what a customer actually did.
A quick tell: count segments that have not been used in a campaign in the last 90 days. A long list of unused segments usually means someone built a structure once and nobody has revisited the logic since.
Second tell: look for a single "engaged" segment used for everything. One engagement definition applied to every send means nobody has thought about the difference between who should get a product launch and who should get a winback attempt.
Check 5: Attribution settings
Check your attribution window before you compare any number to any benchmark. Klaviyo's default attributes a conversion to an email opened or clicked within a set window before purchase, and two accounts with identical performance will report very different revenue if one uses a longer window than the other.
This matters most when an agency shows you a competitor's results. Ask what window produced the number, and whether it counts opens or only clicks. If they do not know, the comparison is meaningless. The same applies to any before-and-after they show you from their own client work.
What a professional audit adds beyond this
Being honest about the limits of the 30-minute version: a real audit adds deliverability forensics (placement testing by mailbox provider, Postmaster Tools review, blocklist checks), template and accessibility review, integration health across every connected app, and revenue modeling of the gaps. If your self-audit found problems, those are the areas where paying for depth makes sense.
What to do with what you find
Write down the three biggest problems you found, then say nothing about them on the call. Let the agency run their audit and see whether their list matches yours.
If they surface things you missed, that is exactly what you are paying for. If they miss what you found, or if their findings are generic enough to apply to any account, you have learned something more useful than any case study.
Frequently asked questions
How long does a proper Klaviyo audit take?
The version above takes about 30 minutes. A full professional audit that includes deliverability forensics, template review, and integration checks takes considerably longer, which is part of what you are buying.
What is the single most common issue in a Klaviyo account?
Missing or incorrect exclusions. Specifically, purchasers who never exit acquisition flows and subscribers who receive messaging written for one-time buyers.
What flow revenue share should we expect?
Klaviyo's benchmark across 183,000+ brands has flows at roughly 41% of email revenue. Treat a large shortfall against that as your opportunity estimate, not as a target to hit in a month.
Should engagement segments use opens?
No. Apple MPP inflates opens 15 to 35%. Build engagement on clicks, site activity, and orders.
Should I give an agency account access before signing?
Read-only access is reasonable and normal for an audit. Full admin access before a contract is not.
Sources
- Klaviyo. The complete checklist to audit your flows.
- Klaviyo Academy. Audit your Klaviyo account.
- Klaviyo. Email marketing benchmarks, 183,000+ brands.
- Klaviyo Help Center. How to reduce spam complaint rates.
- Klaviyo Help Center. Identifying Apple MPP opens.
- Mailgun. Google and Yahoo bulk sender requirements.