A personalization playbook is the list of plays you will run on your website: for each one, which visitors it targets, how you recognize them, what changes on the page, and how you will know it worked. This template turns that into a single table you can review in a weekly meeting. Copy it below, or download the template as CSV.
It sits between two other templates. The personalization audit checklist tells you whether you are ready to personalize. The personalization ROI scorecard reports results once plays are live. This playbook is the middle step: deciding what to run and in what order.
Why a Playbook Beats a List of Ideas
Most personalization programs stall after the first two or three experiences. The reason is rarely technology. It is that nobody wrote down which segment each play serves, so new plays overlap old ones, nobody can say which one drove a result, and the program drifts into a pile of rules nobody dares to switch off.
The fix is a one-row-per-play format where every row must name a signal you can actually detect and a metric with a holdout. If a play cannot fill in both, it is not ready.
The Playbook Table
| Field | What to write | Example of a weak entry | Example of a strong entry |
|---|---|---|---|
| Play ID and name | Short, unique name | "Healthcare" | "P03 Healthcare hero and proof on homepage" |
| Segment | Who the play targets, defined so two people would agree on membership | "Big healthcare companies" | "Hospitals and health systems, 1,000+ employees" |
| Signal | How you detect the segment in real time | "We know they're healthcare" | "Reverse IP industry = Hospitals and Health Care, or utm_campaign contains 'hc'" |
| Expected reach | Share of monthly visitors who will match | "Lots" | "About 600 identified visits a month" |
| Page and element | Exact URL and element that changes | "Homepage" | "/ hero headline, subhead and logo bar" |
| Default | What everyone else sees | (blank) | "Current generic hero" |
| Variant content | The copy, proof or offer shown | "Healthcare messaging" | "Headline about audit-ready patient data; three hospital logos; HIPAA note" |
| Hypothesis | Why this segment should respond | "Personalization works" | "Healthcare buyers bounce when they see no peers and no compliance mention" |
| Primary metric | One metric for the play | "Engagement" | "Demo requests per identified healthcare visitor" |
| Holdout | Share of the segment that keeps seeing the default | (blank) | "10% of matched visitors" |
| Owner | One person | "Marketing" | "Content lead" |
| Review date | When you decide to keep, change or retire | (blank) | "8 weeks after launch" |
| Status | Idea, ready, live, kept, retired | "Live" |
Plays Worth Starting With
These are starting rows for a typical B2B software site. Each uses a signal that is available on most stacks. Replace the content with your own.
| Play | Segment | Signal | Change | Primary metric |
|---|---|---|---|---|
| Industry hero | Top 2-3 target industries | Reverse IP industry | Homepage headline, logos, case study | Demo requests per visitor in segment |
| Company size pricing | Enterprise vs SMB | Employee count from enrichment | Pricing page plan order and CTA ("Talk to sales" vs "Start now") | Pricing page to demo or signup rate |
| Ad message match | Paid search visitors by ad group | utm_campaign or utm_content | Landing page headline repeats the ad promise | Landing page conversion rate |
| Target account welcome | Named ABM accounts | Account list match on reverse IP | Account-specific banner and relevant case study | Meetings booked with target accounts |
| Returning evaluator | Visitors with 3+ sessions who viewed pricing | First-party behavior | Swap "Learn more" CTAs for a meeting or comparison offer | Demo requests from returning visitors |
| Customer suppression | Existing customers | CRM account status | Replace acquisition CTAs with support, upgrade or community links | Fewer customer submissions on sales forms |
For deeper guidance on each type, see industry-specific website personalization, personalizing by company size and pricing page personalization.
How to Run the Playbook
- Fill in reach before anything else. A play that matches 40 visitors a month will never produce a readable result. Sort plays by expected reach and start with the top two.
- Launch one play per page at a time. Two plays changing the same hero for overlapping segments make both results unreadable. If segments overlap, write a priority order (for example: target account beats industry beats company size).
- Keep the holdout on. Without it, any lift is confounded with seasonality and campaign changes. Our guide to measuring personalization ROI explains how to size it.
- Review on the date in the row. Keep plays that beat their holdout, rewrite the content of plays that do not, and retire plays after two failed versions.
- Add one new play per month. A steady cadence beats a big launch of ten experiences nobody can maintain.
Illustrative Example
Illustrative, not real client data. A B2B software company selling to logistics and manufacturing fills in its first quarter:
- P01 Industry hero (live): logistics and manufacturing visitors identified by reverse IP see an industry headline and two peer logos. Expected reach about 1,100 identified visits a month. 10% holdout. Review at week 8.
- P02 Ad message match (live): three paid search ad groups each get a landing page headline that repeats the ad. Reach is all paid traffic, so it reads fastest.
- P03 Customer suppression (ready): existing customers stop seeing "Book a demo" in the header. Measured by the share of sales-form submissions that come from customers.
- P04 Target account welcome (idea): blocked until the account list is synced from the CRM. The signal column makes the blocker obvious.
Frequently Asked Questions
What is a personalization playbook?
A documented set of personalization plays, each defining the target segment, the signal used to detect it, the content change, the metric and the holdout. It turns personalization from a set of one-off experiments into a program you can manage and report on.
How many plays should a B2B site start with?
Two or three. Start with the plays that have the highest reach and the simplest signal, usually industry or ad-campaign matching. Add more only after the first ones have a result you can read.
How is this different from the personalization audit checklist?
The audit checklist tests your readiness: data, content, tools and measurement. The playbook is the plan you run once you are ready. If more than a third of the audit fails, fix those gaps before filling in plays.
Do I need a personalization tool to use this template?
No. Ad message match and customer suppression can run on separate landing pages or simple CMS rules. Plays that rely on company identification need a visitor identification source and a way to swap content, which is what personalization tools provide.