First 90 Days With a Growth Agency: What Should Actually Happen

Direct answer: In the first 90 days with a growth agency, you should see access granted on accounts you still own, a written baseline, a 90-day plan, leak-fixes in month one, a live system in month two, and a keep/kill/change review in month three. If month one is only decks and month three is still "getting set up," onboarding is already slipping.

This is what you should expect from a marketing agency after you sign, especially on ecommerce retainers. It lowers the risk of the strategy call, because you can tell in 90 days whether you bought a team or a slide deck.

It is not a hiring guide. If you still need to pick a shop, how to hire a marketing agency is the evaluation process. Red flags is how to walk away. Agency vs freelancer vs in-house is the operating-model call. Klaviyo agency cost is the pricing math. Best Klaviyo agencies is the shortlist. This post starts after you sign.

Last reviewed: November 9, 2026.

What should happen during agency onboarding for ecommerce?

Onboarding is not a welcome email and a Notion page. It is the period where they learn your accounts well enough to stop guessing, you learn how they work, and both sides write down what "good" means for the next 90 days.

If that sounds slow, it is slower than blasting a campaign on day three. It is faster than spending a quarter on work you cannot measure.

Here is the sequence I want to see:

1. Access, on your accounts.
2. Kickoff, with a named team and a written 90-day plan.
3. Audit and baseline, in dollars, not vibes.
4. First 30 days: leak-fixes and tracking.
5. Days 31-60: the build.
6. Days 61-90: the system runs, then you review it.

We have run this sequence across 150+ clients as an agency. The details change by account. The order does not.

Access: they work in your house

The brand keeps ownership. The agency gets user access. That is the whole rule.

Shopify. Use a collaborator account for a Shopify Partner. You approve the request, you pick the permissions, you can remove them. Collaborators do not count against the store's user limit, and you cannot transfer store ownership to them. Shopify also expires collaborator access after 90 days with no login, which is a useful reminder to keep the relationship active or cut it.

Klaviyo. Add them as a user on your account. Owners and Admins can invite users and set roles. An account has one Owner, and that Owner can close the account. Admins have full access but cannot fully close it. Do not transfer Owner to the agency. Manager is enough for a lot of execution work. Admin is a trust decision, not a default.

Google Ads and Meta. Invite them into your ad account and your Business Manager. Do not move the account into their business. Do not change the login email to theirs. If you ever fire them, you should still be able to log in.

Analytics. GA4, Google Search Console, and whatever pixel/CAPI setup you already have. If they cannot see the same conversion events you see, they will report on a different version of reality.

Give them what the scope needs. Do not give them billing, bank, or "manage users" unless you have a reason. If they cannot explain why they need a permission, they do not need it.

Kickoff: names, constraints, calendar

A kickoff that is only a brand-story slide is a sales call with a new label. The useful version is operational.

You should leave it with:

• Who actually works the account (not who sold it)
• Where you talk (Slack, email, both) and how fast they reply
• What is in scope for 90 days, and what is parked
• Promo calendar, product launches, inventory constraints
• Brand voice: what you will never say, what you always say
• A written 90-day plan with owners and dates

If they cannot name the strategist and the person who will build the work, you hired a logo.

What does a real audit and baseline look like?

The audit is not a 40-page PDF of screenshots. It is a baseline you can compare against on day 90.

Pull the last 90 days of performance before anyone "optimizes" anything. Otherwise you will argue about whether the work worked.

For most ecommerce retainers, the baseline should cover:

Revenue. Total store revenue. Channel-attributed revenue for email/SMS, paid, and organic if those are in scope. Flow vs campaign split on email. That split matters because a pretty campaign calendar with dead flows is a common way to look busy.

Efficiency. Blended marketing efficiency (revenue vs total ad spend) if paid is in scope. Revenue per recipient on email, not just open rate. Cost per acquisition only if you can say whether it is new-customer or blended. Mixing those two is how reports lie.

Tracking. Do purchase events fire? Do UTMs exist? Does Klaviyo's placed-order metric match Shopify in a way you can live with? If tracking is wrong, every later "win" is a guess.

Leaks. Broken flows, suppressed buyers still in sendable, purchaser exclusions missing on acquisition ads, a welcome series that is one coupon and a goodbye. These are the first-30-days jobs.

Constraints. Deliverability, list quality, creative assets, site speed, offer rules. Strategy that ignores these is fan fiction.

The output should be a short written baseline: here is where you are, here is what is broken, here is the 90-day plan, here is how we will know it moved. If you want a second pair of eyes on the email side before that plan is locked, the free Klaviyo audit is the fastest way to get a read.

A US pet accessories brand we took over had demand already. Handmade product, real reviews, customers who actually liked the dogs in the photos. Email was a leftover: one welcome coupon, no segments, templates nobody would send to a friend. The first month was access, audit, and a baseline. We did not "scale email." We documented that the channel was not doing any of the work. That diagnosis is the whole point of days 1-30.

What should the first 30 days with a growth agency produce?

Month one is not the month you get a new revenue graph to post in Slack. Month one is the month you stop flying blind.

By day 30, you should have:

Access complete. Every login that was on the kickoff list is live, or there is a written blocker with an owner and a date.

Baseline delivered. Not "we're still pulling numbers." A document you can open.

Tracking checked. Test purchases, pixels, CAPI, UTM conventions. If they started sending or spending before this, they are optimizing noise.

Leak-fixes. The obvious broken things get touched. A flow that never went live. A suppression list that still includes buyers. An ad campaign that retargets people who ordered yesterday. These are not "the strategy." They are why the strategy can work later.

The 90-day plan, agreed. Channels, sequence, what "done" means at day 90. Account-specific. No blanket "we usually 3x this."

You should also have the first report, and it should look almost boring: here is the baseline, here is what we fixed, here is what we will build next. A report full of early percentage lifts in week two is a warning sign, not a win.

On email and SMS, that usually means the flow audit and the sendable-list hygiene before a campaign calendar. On retention, it means you agree what repeat purchase even is for this catalog before anyone builds a VIP tier. Paid and SEO follow the same logic: fix measurement, then build.

We did not invent a 90-day multiplier for that pet accessories account. Email revenue tripled in two months and the channel hit 24x ROI after the rebuild. That was a brand with demand and a dead channel, not a formula. Your account will not move on the same calendar. Anyone who says otherwise is selling you a screenshot.

If you want us to walk a 90-day onboarding plan against your actual accounts, book a strategy call. Bring the last 90 days of channel revenue and a list of who currently has admin. We will tell you what kickoff should look like, including if we are not the right fit.

What should happen in days 31 to 60?

This is the build month. If month two is still "aligning on goals," they used month one as a second sales cycle.

What should go live depends on what you hired. The pattern is the same: highest-leverage work first, tests with a hypothesis, no drive-by redesign of the whole brand.

Email/SMS. Core flows actually live, not sitting in Draft. Welcome, abandoned checkout, post-purchase, at least one winback path. Campaign cadence that matches inventory and the calendar you already gave them, sent to segments, not the whole list because it is easier. If SMS is in scope, consent and quiet hours get treated as real constraints, not a toggle.

Paid. Account structure that matches how you sell, conversion tracking verified, purchaser exclusions on acquisition, creative tests that change one thing at a time. Weekly notes that mention dollars, not "learning phase" as a personality.

Site/CRO. If it is in scope, one or two tests on pages that already get traffic. Not a full redesign in month two unless that is the project.

You should see artifacts, not just meetings. Flows you can click. Campaigns that sent. Ad sets that exist. A shared tracker of what is live vs draft.

Communication in this window should get quieter and more useful. You should not need a 60-minute call to find out whether the welcome flow is on.

What should be true by days 61 to 90?

By day 90 the core system should be running, not still being "architected."

That does not mean the channel hit a promised number. It means:

• The work from the 90-day plan is live, or the misses are written down with a reason
• You have at least a few weeks of the new system, not just the old one plus a new logo in the footer
• Reporting compares to the day-30 baseline, not to a goal they invented in week 11
• You can answer "what did we learn?" in a paragraph
• You have a next-quarter plan: keep, kill, change

Day 90 is a review, not a birthday. Keep what is working. Kill what was a bad idea. Change what needs another cycle. If they want to "wait one more month before we look at numbers," they are scared of the baseline they should have written on day 30.

For some brands, 90 days is when email starts to look like a channel. For others, it is when you finally trust tracking enough to spend. A US pet bed brand's Klaviyo revenue growing 65.6% to $893K in 2025 was a year of a system compounding, not a kickoff miracle. A Brooklyn spirits retailer's +297% YoY email revenue in Q1 2024 came from a segmentation plan, not from onboarding theater. Do not grade day 90 as if it were year two.

The 90-day roadmap checklist

Use this as the shared punch list. Tick it off with your agency. Progress saves in this browser.

What should reporting and communication look like?

You hired a growth agency to move revenue. Reporting that cannot say that in English is not reporting.

Cadence. Weekly async (Loom, written note, shared doc: pick one and stick to it). Monthly strategy call. Ad hoc when something breaks. If the only time you hear from them is the invoice, they are not in the account.

Metrics. Channel revenue, contribution to store revenue, efficiency (RPR on email, blended efficiency on paid), what went live, what we will test next. Opens and impressions can sit in an appendix. They should not be the headline.

Access to the source. You should be able to open Klaviyo, Ads Manager, or GA4 and see the same story. A PDF that cannot be rebuilt from the accounts is a brochure.

Month-one vs month-three. Month one explains the baseline and the leaks. Month two shows the build. Month three compares to the baseline and recommends the next quarter. If every monthly report looks like a recap of activity, they are billing hours, not owning an outcome.

I will not pretend there is one template that fits every retainer. Email-led work and paid-led work need different views. What they share is this: dollars, against a baseline, on a calendar you both agreed to.

What does the brand still own?

Agencies that "take it all off your plate" are often taking the parts they should not.

You still own:

• The business. Product, pricing, inventory, shipping, CX. Marketing cannot invent stock you do not have.
• The accounts. Shopify, Klaviyo, ads, analytics, domain, billing. They are users. You are the owner.
• Brand voice and offers. They can draft. You sign off on claims, discounts, and anything that will hit customer service.
• The promo calendar. They should not find out about a sale from the storefront.
• Approvals on major changes. New offering, new audience, new send-from domain, budget jumps. Those are yours.
• Anything not in the SOW. If SEO is not in scope, they should not "just start a blog." If CRO is not in scope, they should not rebuild checkout.

Your job in the first 90 days is not to project-manage every email. It is to grant access fast, answer questions once, and show up to the reviews. Slow access is the most common way brands sabotage a retainer they are already paying for.

GOSH email and SMS work starts at $1,500/mo. Retention starts at $2,500/mo. Those figures are on pricing. If a line item is not published there, we quote it after we see the account.

What are the warning signs that onboarding is slipping?

You do not need to wait until day 90 to notice this.

Access is still incomplete at day 14. Someone is "waiting on the collaborator code" and nobody owns the follow-up. Work cannot start. The clock already started.

No written baseline. If they cannot say what the last 90 days looked like in dollars, they cannot say what changed.

They want the logins, not user access. New Klaviyo account under their email. Ad account inside their Business Manager. Shopify Partner that asks you to transfer ownership. Leave.

Reports are vanity. Impressions, followers, "reach." No channel revenue. No comparison to the baseline.

The plan changes every week. A living plan is fine. A new strategy every Monday is how you get to day 90 with nothing live.

You cannot name the people. Sales did the kickoff. A rotating junior is in Slack. The strategist is "the team."

They spent or sent before tracking was verified. That is how you buy a story you cannot audit.

They guarantee a result. Rankings, ROAS, a 3x on email. Nobody controls every variable. A good agency will talk ranges, leading indicators, and what they will do if the plan is wrong.

One of these is a conversation. Two or three is you paying for onboarding that is not happening. Reset the plan in writing. If they cannot do that, you already know.

What to do next

If you just hired someone, print the checklist, put dates on it, and send it to them. See whether they treat it as the job or as an insult.

If you have not hired yet, use the hiring and red-flag posts first, then come back here and ask every finalist to show you their 90-day onboarding, with a sample baseline. The shop that cannot describe days 1-90 will not magically become clearer after you pay the deposit.

If you want a growth-agency kickoff walked against your accounts, book a strategy call. We will go through access, baseline, and a 90-day plan. GOSH Digital is a Klaviyo Gold Partner, founded in 2021 and headquartered in Dubai. 150+ clients, $70M+ revenue driven, $23M+ ad spend managed, 100+ flows built. No pitch deck. Bring the last 90 days of numbers.

---

Related reading:

• How to Hire a Marketing Agency Without Getting Burned
• 9 Red Flags When Hiring a Marketing Agency
• Klaviyo Agency vs Freelancer vs In-House
• How Much Does a Klaviyo Agency Cost?
• Best Klaviyo Agencies for eCommerce Brands