# Test plan — Real estate

The test plan with three columns a property campaign needs: which listing or market the test runs in, which stage of the sale it runs at, and an explicit note on whether the sample can support a conclusion at all.

Everything else is the general test plan, unchanged. One hypothesis with a because in it, one variable, a named control, a minimum run, a read date set in advance and a decision rule for every outcome.

## How to use it

1. Copy this file, or the CSV version if you would rather work in a sheet.
2. Write the hypothesis before the variant, and make sure it contains a because.
3. Name one variable. If there are two, that is two tests.
4. Set the primary metric, the minimum run and the read date before anything is built.
5. Write the decision rule for every outcome, including the one where nothing moves.
6. Leave Result blank until the read date, and then fill it in whatever it says.
7. Keep the finished rows. A year of them is worth more than any single test in it.

## The columns

**Test ID** (Text). A short stable identifier that travels onto the creative file and the campaign names of every listing in the test, because the unit here is a corridor rather than one property.

**Hypothesis** (Text). What you believe and why, with a because in it, and here the reason is usually about the buyer in that corridor: the floorplan will raise attendance because these buyers are upsizing families.

**Variable** (Text). The single thing being changed, held identical across every listing in the test. A variable that changes with each property is not a variable, it is a set of one-off campaigns.

**Control** (Text). What the variant is compared against, named exactly: the way the office would otherwise have run these listings. Both arms must be versions you would have been willing to run for the vendor.

**Variant** (Text). What is being tried, described so somebody could build it from the assets each listing actually has, because a variant requiring drone footage cannot run on a listing that has none.

**Audience** (Text). Who sees the test, identical on both sides. Corridor radius and life stage do most of the targeting here, so the two arms differ in the creative rather than in the audience.

**Where it runs** (Text). The platform, campaign and placement the test lives in, applied the same way across every listing in the test so the comparison holds.

**Primary metric** (Text). The one number this test is decided on, which should be an inspection attended or a registered buyer. Enquiry counts move easily and predict very little about a sale.

**Minimum run** (Text). The shortest run before anybody looks, expressed as a number of listings rather than a number of weeks, because one property produces too little of anything to decide a question.

**Read date** (Date (YYYY-MM-DD)). The day the result is read, set in advance and outside a holiday period, when attendance is dominated by who happened to be in town.

**Decision rule** (Text). What each outcome means, written before the test starts, with the sample caution attached: at these volumes only a large difference should be treated as a result.

**Owner** (Text). The person who builds the test, reads it on the read date, and gets the agents involved to agree to it before any vendor's campaign carries an arm.

**Result** (Text). What happened and what was decided, written on the read date, including the listings whose campaigns were cut short by an early sale. Blank until then.

**Listing or market** (Text). Whether this test runs on one listing or across a market. In property the market is usually the only unit with enough volume to decide anything, so most rows should name one.

**Stage of sale** (Text). Where in the campaign the test runs: pre-launch, launch weekend, open inspections, pre-auction. A variant that wins at launch says little about the pre-auction week.

**Sample caution** (Text). An honest note on whether the volume can support a conclusion, written before the test starts. A row that says the sample is too small is more useful than one that pretends otherwise.

## The test plan

| Test ID | Hypothesis | Variable | Control | Variant | Audience | Where it runs | Primary metric | Minimum run | Read date | Decision rule | Owner | Result | Listing or market | Stage of sale | Sample caution |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|   |   |   |   |   |   |   |   |   |   |   |   |   |   |   |   |
|   |   |   |   |   |   |   |   |   |   |   |   |   |   |   |   |
|   |   |   |   |   |   |   |   |   |   |   |   |   |   |   |   |
|   |   |   |   |   |   |   |   |   |   |   |   |   |   |   |   |

## Before you commit it

- Every row has a hypothesis with a because in it.
- Every row names whether it runs on one listing or across a market.
- Every row names the stage of sale the test runs at.
- The sample caution is written before the test starts, honestly.
- The primary metric is an inspection attended or a registered buyer, not an enquiry.
- The decision rule covers the outcome where nothing moves.
- Every finished row has a result written on its read date.