The packages are named, the pricing holds, the run-sheet exists. Now five real businesses have to sign, get results, and be willing to say so. These first accounts matter beyond their invoices. They decide what your agency is known for, they test every process you wrote, and, handled well, they become the case studies that win the next twenty. This lesson turns the earlier chapters into one sequence for those five accounts.
Pick clients you can move, in a niche you can repeat
Early clients set your trajectory twice over. The work you do for them becomes the work you're known for, and the results you get them become the proof you sell with. So the first five are a choice, not whoever replies first.
The first filter is winnable. A good first client has blockers you can actually fix, a person who can implement what you specify (a developer, or a CMS your team gets access to), and a market where movement is plausible. Think local or specialist, not a head-on contest with national brands for their core terms. The audit tells you most of this before you sign, which is one more reason to sell with it. The audit as a sales tool covers how.
The second filter is repeatable. When client two is also a clinic, or also a store, the keyword process from client one transfers almost whole (Keyword strategy per client), and you're already holding the right playbook (Local and e-commerce playbooks). Unrelated clients are separate educations, paid for in your evenings.
Walk away from two kinds of prospect. The first is anyone who asks you to guarantee a position. Google's own "Do you need an SEO?" guidance warns against firms that promise rankings, and you shouldn't become one. The second is anyone whose site you won't be allowed to change. An engagement without implementation is a report subscription, and it churns the moment budgets tighten.
Run the same sequence five times
The temptation is to treat each client as special. Resist it. The point of the first five is finding out whether your process survives contact with real accounts, and you only learn that by running the identical sequence every time.
The sequence is the earlier chapters in order:
- Win the account with a short, prioritised audit (The audit as a sales tool).
- Onboard in five working days: access, baseline, plan, kickoff (Onboarding a client in a week).
- Build the keyword map sized to their budget (Keyword strategy per client).
- Plan a quarter of content the client will approve (Content calendars that ship), run every piece through the pipeline and its gates (The content production pipeline, Quality gates before publishing), and confirm each page is actually live (Publishing to a client's CMS).
- Report monthly in plain words (Reports clients actually read) and hold the review call on a fixed cadence (The monthly review cadence).
Keep a deviation log. Whenever you do something off-script for a client, write down what and why. Each entry is either a missing step that belongs in the run-sheet, or out-of-scope work that belongs in an add-on (Upsells that help clients). The log becomes the operations manual you'll need later, and each entry takes about a minute.
Over-deliver on process, not on scope
The instinct with early clients is to over-deliver by doing free extra work. It's expensive twice. It resets what the client believes the retainer includes, and it steals hours from your other accounts. The scope you wrote in Scoping and pricing SEO holds from client one, because you will never have an easier moment to establish that it does.
Over-deliver on things that cost discipline rather than hours. The report lands on the agreed day, every month. A ranking drop reaches the client from you, with a diagnosis and a plan, before they've noticed it themselves (Handling ranking drops). Findings arrive as sentences a founder can repeat to their business partner (Presenting findings without jargon). None of this shows on an invoice. All of it is what a happy client describes when a peer asks who does their SEO, and referrals from the first five are how client six arrives without a sales call.
Turn each one into a case study
A case study has four parts: a recorded starting point, the work you did, what changed, and the client's permission to say so. The first three fall out of the process if you run it. The baseline from onboarding week is the before, the plan is the work, and the reporting from Proving ROI honestly is the after, with causation claimed only where you can defend it.
Permission is the part agencies leave too late. Ask at the moment of a win, while the client is looking at a result they're pleased with, and ask for the referral or the public review in the same conversation. Use the client's real numbers with their sign-off, or use none. An unverifiable claim reads as invented, because it usually is.
Frequently asked questions
Can I do SEO on my own?
Yes, and for your own site you should start there. Selling it to clients is different: you need a repeatable process, so the sequence above matters more than any single trick. Run it yourself on a handful of accounts before you hire anyone.
Is SEO dead or evolving?
The work is changing, but clients still need pages that can be found, trusted and acted on. Tools now handle more of the drafting, so your value shifts toward choosing the right keywords, fixing blockers, and proving results honestly. Those are the parts this playbook builds your reputation on.
How to get clients for website business?
Sell with a short audit that shows a prospect their own fixable problems, and pick people who can implement what you recommend. Then let early results do the selling: ask each happy client for a referral or public review at the moment of a win. Referrals from your first accounts are how later clients arrive without a sales call.
Can ChatGPT do SEO?
It can speed up parts of the job, such as drafting or organising a keyword list. It can't get access to a client's CMS, confirm a page is live, or take responsibility for a ranking drop. Keep your pipeline and quality gates in place so a person checks everything before it publishes.
What to take away
- The first five clients decide your niche and produce your proof, so choose them for winnability and repeatability instead of taking whoever signs first.
- Run the identical sequence on all five and log every deviation: each entry is either missing process or missing scope.
- Over-deliver on discipline (reports on the agreed day, early warnings, plain language), never on free extra work.
- A case study is a baseline, the work, the result and permission, and the moment to ask for permission is the moment of a win.
Next
Scaling to fifty sites covers what has to change when the playbook that ran five accounts has to run fifty.