Client readiness: the gauntlet, part one

This is the final exam, and it looks nothing like a quiz. Below are four clients. Each one is invented, but nothing about their problems is: every scenario is built backwards from a failure class our checks catch on real domains, the kind that lands in a practitioner's inbox as a confused paragraph and a plea. Each scenario runs three decisions deep, and the situation evolves with your answers, the way a real engagement does.

How to play it honestly. Commit to an option before you tap anything, the way you'd have to commit on a client call. No rewinding: your first instinct is the data point. And after each decision, read the feedback on every option, including the ones you didn't pick. That's where the value is. The wrong answers here aren't filler; they are the things well-meaning people genuinely do (resend the campaign, buy the warm-up service, mail the apology), and each one carries a consequence worth knowing before a client pays you to know it.

One more thing before you start. You have everything you need. Every decision below leans on a lesson you've already done: reading bounces, reading headers, the fix order, the triage tree, who owns what on shared infrastructure. If a step feels unfamiliar rather than merely hard, that's a signal to revisit the part it came from, not to guess harder.

Keep count of your best-call decisions as you go. There are twelve. Scoring guidance waits at the end.

Client one: the florist

A two-person flower shop. Day-to-day mail on Google Workspace, weekly campaign to 6,000 subscribers through Mailchimp. They found you through a friend and they are worried.

The open rates that halved in March

The owner shows you the Mailchimp reports: open rates ran near 40% for two years, then halved in early March and stayed down. Nothing about the newsletter changed. The only event she can name: in late February they switched website hosting companies, and the new host 'handled the whole transfer'.

What do you do first?

Go through the March campaigns looking for subject lines or content that could have tripped filters.Check the domain's DNS: SPF, DKIM selectors, DMARC, against what should be there.Have her resend the March campaigns to everyone who didn't open them.Open a ticket with Mailchimp asking why deliverability dropped.

Client two: the consultancy

A five-person engineering consultancy on Microsoft 365. Proposals go out through an e-signature tool that sends from their domain. New business depends on those proposals arriving.

The proposals that vanish

The managing partner tells you proposals to their biggest prospect, a manufacturer, get no reply. His contact there swears nothing arrived. Ordinary one-to-one Outlook mail to the same person arrives fine. There are no bounces anywhere. Proposals to other clients seem to land normally.

What's your first move?

Treat it as a reputation crisis: blocklist checks, full audit, the works.Map the pattern precisely: which path fails (the e-signature tool), which succeeds (Outlook), and to which recipients.Resend the proposal from the partner's personal Gmail so it gets through.Ask the contact at the manufacturer to check his spam folder.

Client three: the nonprofit

A small arts charity with a 9,000-address house list and years of clean sending through their ESP. Their annual gala, the fundraising event of the year, is five weeks out.

The borrowed gala list

Two weeks ago a partner organization shared its 5,000-address list as a gala favor. The charity imported it and mailed everyone once. Since then: a wave of unsubscribes, some angry replies, and a complaint uptick in the ESP dashboard. The director wants the next three gala emails to go to the full combined list, on schedule.

What do you do first?

Send the next email on schedule. It's a good cause and the gala date won't move.Quantify before deciding: the spam-rate chart in Google Postmaster Tools, the ESP's complaint numbers, and the imported addresses split into their own segment so the two populations can be treated differently.Run the imported list through an email verification service and keep whatever passes.Send an apology email to the imported addresses.

Client four: the store

An online kitchenware store, mid-November, heading into its biggest weeks. It sends through a mainstream ESP on shared IPs, with a consented list and clean history.

Listed in peak season

The owner calls in a panic. A free 'blacklist scanner' says the domain is listed on six blacklists, and this morning a handful of campaign sends to business and small-provider addresses bounced with messages that name a blocklist. Gmail placement looks normal so far.

What's your first move?

Start delisting requests with all six lists today.Separate evidence from noise: read the actual bounce messages, which name the list and the listed entity, and establish whether the listing is on the store's domain or on the ESP's shared sending IP.Reassure the owner: Gmail looks fine, and Gmail is most of the list.Migrate to a new ESP immediately so the store has clean IPs before the peak.

Scoring yourself

Twelve decisions. Count your best calls.

Ten or more, with reasons you could say out loud to a client: strong. That's the judgment the practitioner track has been building toward, and the graduation lesson's checklist will feel like confirmation rather than news. Seven to nine: solid instincts with specific gaps; the feedback you read on the misses names which part to revisit, and it's usually one part, not five. Below seven: no verdict on you, but a clear one on sequence. Go back through the misses and notice whether they cluster (authentication evidence, list judgment, ownership of shared infrastructure) before booking anything with a paying client.

One caution on a perfect score: twelve out of twelve reached by pattern-matching the shape of the options is worth less than nine reached by reasoning from evidence. These four clients told you the truth. Their symptoms pointed, more or less honestly, at their causes.

That courtesy ends now. Part two of the gauntlet is the harder half: symptoms that point at the wrong culprit, damage the client inflicted themselves last Tuesday and won't mention, and at least one case where the professional answer is that this isn't a deliverability problem at all.