# How to run email marketing for clients and keep them past month six

> The weekly rhythm, onboarding, monthly approvals, team roles, AI use and reporting that keep an agency's email clients after the flows are built.

By The Wizel Team · Published 2026-11-19

An email agency can do good work and still be let go a few months in, once the flows are built and the client decides campaigns plus good-enough flows are something they could run themselves. The work hasn't stopped; it has gone out of sight. This chapter is the operating rhythm that keeps it in sight across a book of clients.

**Month 4-6** when many email retainers end: the flows are built and the agency's work goes quiet

The fix is to work as a partner, not an executor: know the brand's numbers better than the brand does, bring ideas unasked, and sit in strategy with the person who decides.

## Show the work every week, on a fixed rhythm

Most quiet churn is a communication failure. The client pays for work they can't see, and silence reads as no work. Keep a set rhythm for each client, hardest in the first two weeks, when buyer's remorse peaks: reply within minutes (an ETA counts), share a roadmap at signing and land a win in week one.

One agency's standard, as a starting point. Adjust it per client.

| Layer | When | What the client gets |
| --- | --- | --- |
| Replies | Within about 30 minutes in work hours | An answer or an ETA; nothing left open overnight |
| Proactive updates | 3+ per client per week | A standout campaign, an A/B result, a strong number deep in a flow |
| Weekly update | Monday or Tuesday | Revenue, last week's sends, test results, approvals needed, the next two weeks' campaigns confirmed |
| Call recap | After every call | Decisions and action items |
| Monthly report | By about the 7th | Last month, once its attribution window has closed |
| Quarterly review | Each quarter | Before and after against the starting baseline, and the next three months |

Keep the weekly update to one page in the same shape every week, so it becomes a ritual:

- **Last week's win:** one proof point with a number.
- **What changed:** two or three insights, each with its so-what.
- **This week's actions:** two to four, ranked by effort and impact.
- **One new idea:** an experiment nobody asked for.
- **Watchlist:** one or two metrics moving the wrong way.

Steady beats polished: a quick screen recording of a change you just made keeps a client longer than a perfect report. Taper calls as trust builds; one agency runs them weekly in month one, then every two weeks.

## Onboard a new client: audit first, then early wins

Close to kickoff the same day: on a yes, send the contract, the invoice and the onboarding form, and book the kickoff before you hang up. Then:

1. **The onboarding form,** about 30 minutes to fill in: Klaviyo and Shopify access, bestsellers, competitors, copy and design do's and don'ts, customer objections, discount history, and what their last agency did well and could have done better.
2. **The kickoff call,** 20-25 minutes, with the account reviewed and a first calendar in hand. Set reply times, turnaround, revision rounds and approval windows, or the client sets the rhythm for you.
3. **Settle the attribution window.** Most brands stay on Klaviyo's default (5 days, opens and clicks). If you want it tighter, set it from day one or explain the change on a call: the numbers drop, performance doesn't.
4. **Audit before you rewrite.** A 10% welcome offer while the ads promise 65% off is a structural fix, not a copy fix. The [flows revenue check](/blog/klaviyo-flows-revenue-check) shows which flows earn and which are missing.
5. **Stack visible wins.** One agency's timeline: pop-up rebuilt within about 3 days, a first campaign or flow presented on days 7-10, first campaigns live around day 14.

Say upfront that timelines start once access and assets are in, and allow about a week of prep before the first campaign. Run a short, incentivized survey of the client's customers early: it hands the copy their real words and objections.

## Plan each client's month and get approvals in one pass

Endless back and forth comes from approving too late and approving the wrong thing. A fixed monthly cycle solves the first:

1. **Mid-month:** send the client a planning form for next month's sales, launches and events. What goes into the month is in [planning a month of email](/blog/plan-a-month-of-email).
2. **Around the 20th-23rd:** share next month's calendar for the client to approve the plan.
3. **About 10 days before each send:** the client approves the finished email. Never fewer than 3-5 days, and rarely more than 14-21. An approval asked for the day before gets no real look, because nobody wants to delay the send.

For the second, have clients approve the finished email only: approving copy before design adds a round, and copy still changes once designed. Cap revisions at two rounds, or offer unlimited revisions turned around within 24 hours, which rarely costs much: heavy revising just means fewer emails go out.

Write the agreement as an email count with a split, such as 12 of 16 monthly emails to campaigns and 4 to flow tests. Then every request beyond it is a choice: absorb it and say it's free this time, or price it as an extra. On strategy, push back once with reasons, and A/B test their version against yours when the stakes allow.

## Who does what on the team

Build the team in pods, each an account manager, a copywriter, a designer and a platform technician on the same 6-8 clients, with the account manager at the center of each one.

Who owns what, as practitioners describe it.

| Job | Owns | Watch for |
| --- | --- | --- |
| Account manager | Client communication, the brief, revisions, the calendar; often the copy | The riskiest hire. Start a new one on 3 clients, not 8 |
| Copywriter | The words, and the design brief section by section | A separate copywriter adds a handoff; some add one only where copy needs a higher standard |
| Designer | The email, and at volume its slicing and upload | Design is the biggest time sink, so usually the first hire |
| Platform technician | Building and scheduling in Klaviyo | The job AI tools are most likely to take over |
| Founder | Sales, the standard, a weekly quality pass | Relaying messages between client and team makes the founder the bottleneck |

Run one board per client and one card per email, from idea to sent; five clients at three emails a week is already 50+ emails in flight. Every email follows one chain: the account manager's brief, the copywriter, the designer, the account manager approving it for the team, the client, then back to the account manager, who fixes small revisions and sends bigger ones back.

Keep quality control an overlay, not a gate. One founder spends a weekly block of about two hours speed-running the boards, comments only on work below standard and leaves pre-send approval to the account managers. When an account manager takes over a client, introduce them as an addition, never as your replacement.

## Use AI in delivery without the work going generic

AI output reads generic when it's missing context. Let AI draft inside full brand context and a design system people built; keep taste, strategy and the final edit human.

- **One workspace per client,** fed by the brand's site and sample emails. Never let one client's data reach another's.
- **A brand hub the AI reads first:** the onboarding form, the brand's own best practices and a feedback log.
- **Voice from evidence.** Take 3-5 artifacts (the last 90 days of sends, the about page, founder letters, social posts) and distill 5-7 short rules: rhythm, words they use and never use, I or we, openings and sign-offs.
- **Feedback as data, not rules.** At one agency, a single "use a fade" comment made its design tool fade every email. Log what worked with its context instead.

How far to trust the draft is where practitioners split. Some report AI getting about 80% of the way, with a copywriter spending about 10 minutes humanizing each campaign. A stricter rule caps AI at about 30% of a first draft. Either way, put a human edit at every hand-off, and when the team keeps redoing AI output, fix the brief.

Automate checks, not insight. An AI report that flags a broken flow earns its keep; deciding why a campaign worked still needs a person. That judgment, with taste and a real relationship with the client's founder, is what stays defensible as production gets cheaper.

## Report so the client reads it

Lead with insight, not data: observation, insight, action, the most meaningful finding first. Numbers the client can pull from Klaviyo themselves add nothing. The monthly report covers:

- **Email revenue in dollars,** campaigns and flows split, month on month and year on year. Not email's share of sales, which falls whenever another channel grows.
- **Retention:** repeat purchase rate, the LTV trend and net list growth (new subscribers minus unsubscribes). Treat flow revenue as directional; [measuring email after Apple's privacy changes](/blog/measure-email-after-apple-privacy) explains why.
- **Tests and what they taught,** with the weakest campaigns framed as next month's focus.
- **A 30/60/90-day roadmap,** which you then deliver on.

Between reports, the watchlist catches decay before the client sees a revenue dip:

Starting thresholds, not rules. Tune them to each client's own history.

| Signal | Trigger | Action |
| --- | --- | --- |
| Repeat purchase rate | Down more than 3 points month on month | A second-purchase or replenishment flow |
| A flow's placed-order rate | Down more than 15% against its 8-week baseline | Refresh the offer and A/B test it |
| Email revenue | Down in dollars 3 weeks running | Check cadence and segments |
| Sign-up form conversion | Below about 2% | Test a new offer, format or trigger |
| Spam complaint rate | Climbing (keep it well under 0.3%, ideally under 0.1%) | Pull broad sends back to engaged segments |

Don't anchor clients on click or conversion rates, or the dip after peak season reads as failure. Flag a big win before they spot the spike. Pre-empt "email went up, so why didn't total revenue?": 40% attributed to email doesn't mean +40% in total sales. Across your own book, compare rates and trends, not totals, or the biggest list always looks like the best work; [comparing email performance across stores](/blog/compare-email-performance-across-stores) shows how.

## Keep the retainer: score every client every week

Score each client every Monday. Green is on track. Amber is strong delivery with weak communication, or the reverse. Red is missed deadlines, poor communication or churn risk. The churn signals are slower replies, skipped calls, "things are good enough" and questions about budget. When they show up:

1. **Re-anchor on value:** put the rolling six-month win summary in front of them.
2. **Re-enter strategy:** propose a fresh initiative with real upside, a reason to stay rather than a discount.
3. **Only if they open the exit conversation,** offer a smaller scope at a smaller retainer. A downgrade keeps the relationship and the data access, and is far easier to grow back from.

> Bring the idea before the client asks. That one habit is what makes leaving feel risky.

Keep a founder-level relationship even after an account manager runs the work. And accept that not every client is worth keeping: one agency waited a year to end a demanding, low-paying client, who quietly shopped around and left first.

## Price it so you can keep doing it well

Charging enough is what pays for better people and lets you turn down bad fits. The method is in [pricing email campaigns as an agency](/blog/pricing-email-campaigns-agency); the parts that bear on keeping clients:

- **Sell a number of emails, not "everything".** One agency sells tiers of 12, 16, 20 and 24+ emails a month, and clients move up or down a tier any month.
- **Charge at least about 2.5x a client's monthly delivery cost,** and log hours per client to see which retainers lose money.
- **Negotiate deliverables, never price.** If a budget is short, cut scope to fit.
- **Raise prices with a case, on a call,** after three to five months, showing what the client keeps from email after your fee.

When a new client signs, start their program from the top with [how to run modern email marketing](/blog/how-to-run-modern-email-marketing), which sets out what to build first. Your audit shows where the client stands against it.

## Onboard a new email marketing client

1. **Close to kickoff the same day**: On a yes, send the contract, the invoice and the onboarding form, and book the kickoff call before you hang up.
2. **Collect the onboarding form**: Keep it to about 30 minutes: Klaviyo and Shopify access, brand and products, competitors, copy and design do's and don'ts, customer objections, discount history and what the last agency did well or badly.
3. **Run a short kickoff call**: Review the account first, bring a first calendar, and set expectations on reply times, turnaround, revision rounds and approval windows.
4. **Settle the attribution window**: Agree the attribution window at the start, and move an existing client to a tighter one on a call rather than by message.
5. **Audit before you rewrite**: Check offers, flows and deliverability for structural problems before changing any copy.
6. **Land early wins**: Rebuild the pop-up within days, present a first campaign or flow in the second week and get first campaigns live around day 14.

## FAQ

### How do email agencies keep clients past the first few months?

By keeping the work visible after the flows are built. A one-page weekly update with a win, what changed, this week's actions, one new idea and a watchlist, plus a monthly report that leads with insight, keeps the value in front of the client.

### How far ahead should a client approve an email?

About 10 days before it sends is a good working rule, and never fewer than 3-5 days. An approval asked for the day before the send gets no real look, because nobody wants to delay it.

### What should an email agency's client onboarding include?

A short onboarding form (about 30 minutes), a 20-25 minute kickoff call, agreement on the attribution window, an audit of the account before any rewriting and a few visible wins in the first two weeks.

### Who does what on an email agency's team?

An account manager who owns the client and the brief, a copywriter, a designer and someone who builds and schedules in Klaviyo. Practitioners group them in pods of about 6-8 clients each.

### Can an agency use AI to write client emails?

Yes, if the AI works inside full brand context: one workspace per client, a brand hub it reads first and a human edit before anything reaches the client. Without that context the output reads generic.

---

Source: Wizel (https://www.wizel.ai/blog/run-email-marketing-for-clients)
Last updated: 2026-11-19