Agency Problems
Part of the Agency Problems cluster
What Should an Ecommerce Agency Actually Deliver
An ecommerce agency should deliver three concrete things: a documented baseline of your numbers before the work starts, a prioritized list of what is being fixed and why, and results you can verify in your own dashboards — Shopify, GA4, Ads Manager, Klaviyo — not in the agency's reporting layer. Everything else is presentation. Most founders have never seen this standard written down, which is exactly why "we're optimizing" survives as a deliverable. This page defines the standard, so you can hold your current agency to it — or your next one.
Why Does Every Engagement Need a Documented Baseline?
A baseline is a snapshot of the metrics that matter, taken from your own platforms before any work begins: blended CAC, new customer ROAS, conversion rate, AOV, flow revenue percentage, repeat purchase rate — whichever numbers the engagement exists to move. It should be written down, dated, and agreed to by both sides in the first week.
Without a baseline, "we improved performance" is unfalsifiable. The agency compares against whatever period flatters the work; seasonality gets claimed as results when it helps and blamed when it hurts. With a baseline, every monthly conversation has a fixed reference point, and the question "is this working?" has an answer both sides already agreed on how to measure.
The baseline conversation is also the cheapest possible test of an agency before signing. An agency that resists documenting your starting numbers — or wants to measure success only inside its own reporting tool — is telling you in advance how the accountability conversation will go in month six.
An agency that will not write down your starting numbers is telling you in advance how month six will go.
What Does a Real Plan Look Like, as Opposed to a Service List?
A service list says what the agency does: creative refreshes, campaign management, flow optimization, monthly reporting. A plan says what is broken in this specific business, in what order it will be fixed, and why that order — prioritized by how much each gap is costing, not by what the agency finds easiest to staff.
The difference shows up in one question: why this, first? A working agency can answer from your data: "flow revenue is at 9 percent of email revenue against a 25 to 35 percent benchmark, so the welcome and abandoned cart flows come before any campaign calendar." A service-list agency answers from its package tiers.
Every item on a real plan carries a named expected result and a window — "this change should move this metric by roughly this much within 60 to 90 days." That sentence is what makes accountability possible later. If nothing was predicted, nothing can be missed, and the engagement can run on vibes indefinitely.
What Should the Monthly Report Actually Contain?
The outcome metric, against the baseline
Not activity. The number the engagement owns — blended CAC, new customer ROAS, flow revenue percentage — compared to the documented day-one snapshot, from sources you can open yourself.
What changed this month, and why
The specific tests, builds, and changes shipped, each tied to an item on the prioritized plan. Effort without a reason attached is activity reporting.
Expected vs actual, honestly
What each change was predicted to do and what it actually did — including the misses. An agency that never reports a failed test is not testing.
What comes next, and why it is next
The following month's priorities, ranked by expected impact on the outcome metric, so the plan stays a plan instead of decaying into a service list.
The single test that separates real reporting from presentation: every claim should be verifiable in your own dashboard. If you have not yet run that verification on your current agency's last report, the how to know if your agency is working check is the place to start — it takes under 10 minutes per channel.
This standard is also the one InfinitHive holds itself to, by structure rather than promise: results are verified in the client's own Shopify, GA4, or Klaviyo data, and the retainer does not activate until a verified 10 percent monthly revenue lift shows up there. An agency's willingness to be measured in your data, not theirs, is the standard — whoever the agency is.
The Standard at a Glance
Day 1
When the baseline should be documented — before any work starts
4 items
Every monthly report: outcome vs baseline, changes made, results vs expectations, next priorities
100%
Of reported claims should be verifiable in your own dashboards
60–90 days
A stated window attached to every expected result, so 'be patient' has an expiry date
1 metric
Every engagement owns at least one outcome metric — named in writing, not implied
0
Deliverables that should require the agency's own reporting layer to verify
Related Pages
How to Know If Your Ecommerce Agency Is Actually Working
The 10-minute check that tests your current agency against this standard: the outcome metric in your own data, plus whether anyone can explain what moved it.
DecideShould I Fire My Ecommerce Agency?
If your agency falls short of the standard, the next question is whether the gap is fixable execution or a structural wall a new agency would hit too. The decision framework.
Frequently Asked Questions
What to Do Next
If you have never had a documented baseline of your own numbers, that gap is bigger than any single agency decision — it means nobody can currently prove what is working. Having the full revenue system mapped and baselined is the right first step.
Free · 20 minutes · No pitch