- Ask for dates and deliverables, not "month one is discovery." My automations go live in the first seven days.
- Everything an agency builds in your Klaviyo account should belong to you, and stay there if you part ways.
- Be wary of a long lock-in. Results are never guaranteed, so a contract that traps you is a risk you carry alone.
- An agency that leads its report with open rate is showing you the easiest number to inflate. Ask about clicks and revenue per recipient.
- Put who can change attribution settings in writing, because changing them changes how much revenue email gets credit for.
Why not just look at case studies?
Case studies show what went right somewhere else. The questions below cover what goes wrong after you sign: slow delivery, fuzzy reporting, contracts you can't leave and accounts you don't fully control. Ask every agency the same 12 and compare answers, not personalities. If an agency offers a free audit first, here's how to tell a real free Klaviyo audit from a sales pitch.
12 questions to ask an email marketing agency
1. What happens in the first 7, 30 and 90 days?
Why it matters: Slow delivery starts with a vague plan.
Good answer: Dates and deliverables. Here's mine. The automations get deployed in the first seven days. In the first 30 days the frameworks are installed, campaigns and flows are set up, and you understand the thinking behind them. But I'd tell you not to expect drastic results in that first month unless there's a specific reason. In the second 90 days there's real data to work from, including real reorder periods, so that's where I'd expect the bigger move. I go through it step by step in what the first 90 days with an email agency look like.
Red flag: "Month one is discovery," with weeks of calls before anything goes live. Or the opposite: a promise of a huge month one before anyone has seen your data.
2. What access do you need in my Klaviyo account, and why?
Why it matters: The wrong access level stalls onboarding before it starts.
Good answer: A clear role for a clear job. For a look-only audit, an Analyst role is fine, because the auditor only needs to observe. For the actual work, I ask for admin. The most common mistake I see is asking for admin and getting a manager role instead. I've had a brand add me as a manager, which caused a massive delay because I couldn't add a team member to deploy the strategy. Here's how to give an agency access to Klaviyo without that back-and-forth.
Red flag: They want the account to sit under their own login, or they can't explain why they need the access they're asking for.
3. Who owns what you build if we part ways?
Why it matters: If the agency controls the work, leaving means starting over.
Good answer: You do. When a brand leaves me, everything I created stays: automations, segments, popups, all of it. I won't delete any of it. What stays mine is the strategy in my head, not the work in your account. We give you everything we create, and it belongs to you.
Red flag: Flows and templates that "belong to the agency," or a vague answer about what happens at the end. If you're already in that spot, read how to switch email marketing agencies before you give notice.
4. How long is the contract, and what's the notice period?
Why it matters: Getting locked into something that isn't working is how founders get burned.
Good answer: A clear notice period, in writing, before you sign. Mine is 15 or 30 days, depending on the contract and what the brand prefers. I don't want to hold any brand hostage. If you want to leave, I want to get you out as soon as possible.
Red flag: A long lock-in with no way out. That's the contract term that hurts brands most, because results are never guaranteed. It's even worse with long software contracts, where a provider can end up holding your data hostage.
5. If I pay up front, what does that payment buy?
Why it matters: Paying up front isn't the problem. Paying up front for work nobody has defined is.
Good answer: Prepayment can be a good thing, depending on the situation. My 90-day accelerator is laid out from start to finish so a brand knows exactly where it's headed. The test is whether the agency can show you the defined setup your money pays for.
Red flag: Prepaid time with no defined setup behind it. If they can't deliver that properly, you probably shouldn't hire them. For what retainers usually run, see what a Klaviyo agency costs.
6. Who can change the attribution settings?
Why it matters: Klaviyo's attribution settings decide how much revenue email gets credit for. Change the window or what counts, and the "email revenue" number moves without a single extra sale.
Good answer: The settings get agreed before you sign, or written into the contract, along with who's allowed to change them. That way the report you read in month six is measured the same way as the one in month one. I explain the settings in Klaviyo attribution.
Red flag: One big "email revenue" number and no answer to how it's counted, or who set it up that way.
7. What do you report on?
Why it matters: What an agency reports is what it's working toward.
Good answer: Clicks, revenue per recipient, and email's share of store revenue. Across 4 consumable brands I work with, click rate tracked revenue at 0.65. Open rate tracked it at 0.32.
Red flag: Open rate as the headline. In that same data, Apple Mail auto-opens pushed open rates past 85% on some segments. An open rate can look great while revenue goes nowhere. More on this in open rate vs click rate.
8. Do you guarantee results?
Why it matters: Nobody knows what your list will do until they've seen it.
Good answer: No. Results are never guaranteed. A good agency will tell you what it plans to build, how it will measure it, and when you'll know whether it's working.
Red flag: Yes. A guarantee before anyone has opened your account is a sales line.
9. Who does the work on my account day to day?
Why it matters: The person on the sales call isn't always the person building your flows.
Good answer: A name, and ideally a chance to meet that person before you sign. You want to know who's writing, who's building, and who you go to when something breaks.
Red flag: "Our team," with nobody specific attached to your account.
10. What would you build first?
Why it matters: The answer shows whether they'll look at your account or run a template.
Good answer: Check the data first, fix what's broken, then build what's missing. For most consumable brands that means flows before a bigger campaign calendar. Across those 4 brands, replenishment flows earned about $2.23 per person, about 28x the best campaign, and only 2 of the 4 had one. Welcome email #1 earned about $4.50 per person. A good agency should be able to tell you which flow your account is missing.
Red flag: "We'll start with a weekly newsletter" before anyone has checked whether your welcome series and abandonment flows work.
11. How do you decide who gets each campaign?
Why it matters: Blasting everyone looks big and earns small.
Good answer: Segments built on who clicks and buys. At a sports training tech brand, an engaged send to 7,145 people earned $0.19 per person. A blast to 49,475 earned $0.04. Once there's data, I condense campaign send lists: fewer people, more revenue, more qualified leads. Here's how I think about email segmentation for ecommerce.
Red flag: Every campaign goes to the whole list, every time.
12. How do you test, and how do you pick a winner?
Good answer: Each test is judged on the number it actually tests. A subject line test is judged on opens, a content or call-to-action test on clicks, an offer test on revenue, and a segment test on revenue per recipient. And one test isn't a rule: I run a test across about five campaign sends, at normal send sizes, before I treat the result as a rule. Here's how to set up a Klaviyo A/B test.
Red flag: "We tested it once and it won."
What should you do with the answers?
Lay them side by side. The agencies worth hiring give specific answers: dates, roles, numbers, and who owns what. The ones to skip get vague exactly where it matters, which is usually ownership, contract length and reporting.
Before you sign with anyone, pull a baseline from your own account: click rate, revenue per recipient, and email's share of store revenue. That's how you judge the work later, instead of taking the report's word for it. And if you already have an agency and the answers don't hold up, here are the signs your email agency isn't working.
What should you do the day you part ways with an agency?
Remove their access. To Klaviyo first, then every other platform they were in. I've had brands keep me in their account for over a year after we stopped working together, and I still don't understand why brands aren't more diligent about it. It takes two minutes.
Where these numbers come from
The pooled numbers come from 4 consumable brands I work with: 2,897 campaigns and $17.7M in email revenue. That's a small sample, so treat them as directional. The segment numbers are from a single brand.
FAQ
What's the biggest email marketing agency red flag?
A results guarantee, or any hesitation about you owning everything built in your account. Either one tells you how the relationship will go when things get hard.
How fast should an agency get flows live?
Ask for a date. My automations are deployed in the first seven days. Brands that are already sending tend to see results faster, because there's more data to show where to take things.
What access should I give an email marketing agency in Klaviyo?
For a look-only audit, an Analyst role is enough. For an agency doing the work, give the access they need to do the whole job. I ask for admin, because a manager role can't add team members.
Should I prepay an email marketing agency?
It can make sense if the money is tied to a defined setup. If an agency can't tell you what the first payment buys, don't send it.
What should a notice period be?
Short enough that you're not trapped. Mine is 15 or 30 days. Anything that locks you in long-term puts all the risk on you.
Founder of Bardown Digital, a Chicago-based Klaviyo email and retention agency for Shopify brands. He's worked with brands doing $500K to $100M a year, with $100M+ in email-attributed revenue across them.