Personalisation projects often start in the wrong place. The targeted hero gets the workshop. The fallback, the page shown when no rule matches or a signal is missing, is whatever was already there.
Treat that fallback as the page most people will see. Design it with the same care as the variant you are proud of. The clever version can wait until the ordinary page stands on its own.
A rule is a guess, and guesses miss
Nielsen Norman Group draws a line between two ways of tailoring a page. Customisation is the visitor choosing. Personalisation is the system choosing. In a 2016 article on the difference, Amy Schade describes the aim as delivering what matches someone’s needs “with no effort from the targeted users.” The cost is in the next breath: “The downside is that we’re at the mercy of the computer’s ability to guess what each user needs.”
A guess has a miss. The visitor is not in a country you wrote a rule for. The campaign link arrives without its parameters. Detection returns nothing. The condition cannot be evaluated. In every one of those cases the page still has to show something, and that something is the default.
Schade’s follow-up, 6 Tips for Successful Personalization, is plain about a bad guess. If someone is tagged as the wrong type, personalisation “can result in a frustrating experience that might annoy the user at every visit, or worse, keep them from using the system altogether.” The same piece says: “Even the best-planned systems can go awry.” People also need a way back to information that was not tailored to them.
The default is that way back. It is also the page for everyone you never wrote a rule for.
Google makes a related point for sites that change by place. Its guide to managing multi-regional and multilingual sites says geotargeting “isn’t an exact science”, so you should plan for people who land on the “wrong” version. It also warns against adapting content from IP location, because that analysis “is difficult and generally not reliable.” Any signal can be missing or mistaken, and the page those people see still has to make sense.
The same guide notes that if you change a page dynamically from language settings, Google might not find every variation, because its crawler usually comes from the USA and does not send a language preference. The version that exists before any guess is the one you can count on being the page.
A broken fallback is the page people remember
Picture a bike shop in Bristol. Ahead of autumn they ask their agency for one personalised block. Visitors from an email about wet-weather kit see a hero for jackets and mudguards. Everyone else should see the ordinary shop: bikes, servicing, a way to book.
The agency builds the email hero with care. The default is whatever the homepage already was: a summer sale that ended in August, and a banner about a bank-holiday closure. It stayed because the interesting work was the rule.
Now count who actually sees the summer sale:
- A rider who opens the site from a bookmark, not the email. No campaign match. Default.
- A parent buying a first bike, who was never on the list. No match. Default.
- A visit where the country cannot be detected. Default.
The jacket hero, the one that took the afternoon, is the exception. The stale sale is the shop.
A default that is empty, half-written or left over from the last campaign is worse than no personalisation at all. Visitors do not know a rule failed. They think this is the page.
Start with a page that does not need the rule
The article that defines personalisation also says what to build first. Personalisation and customisation “should enhance an already good experience, rather than try to fix a poor one.” And:
“it’s the designer’s job to prioritize the ideas and create a tight base-level design that can subsequently be augmented through personalization and customization.”
That is the order of work. The base-level design is the default. The rule augments it. It does not rescue it.
For a page like the bike shop’s, a tight default does four jobs:
- It stands alone. Switch every rule off and the page still says what you sell, who it is for and what to do next. If it only makes sense beside the variant, it is a gap, not a default.
- It is not a previous campaign. An expired offer is not neutral. It is wrong. The unmatched visitor did not ask to see the thing you forgot to take down.
- It does not pretend to know. “Welcome back” for a stranger, or a price for a market you do not serve, is a worse miss than a plain page. Keep the ordinary information available when a rule swaps a block.
- It is the version you review. Looking only at the matched preview means you are checking the exception. Open the page with no campaign parameter, with detection unavailable, and as a visitor you did not write a rule for.
Schade also warns against more segments than you can support with real content. The same limit applies to rules. One well-kept alternative beats four that drift.
A practical order for the first rule
Write the default as if the rule will never fire. Headline, offer, image, call to action. Specific enough to be useful, and general enough to be true for anyone who lands there. For the bike shop: servicing, bikes and kit, and a clear way to book or browse. No season that can expire unnoticed. No city name unless the shop really is only in that city.
Change one part, not the whole page. A hero, a banner or an offer. The jacket block can replace the hero for people from the email. It should not be the only place the page explains the shop. When the rule matches, the rest of the page still has to hold together. When it does not, nothing should collapse into a blank block.
Name the misses before you publish. No rule matches. A condition cannot be evaluated. The visitor is somewhere you did not list. The link lost its parameters. For each one, write what the visitor sees. If the answer is “the old page” or “an empty space”, you do not have a default yet.
Review the default with someone who did not build the rule. Ask whether the page still works, not whether the logic is clever.
Then switch the rule on.
Look at the fallback before anyone else does
Designing the default on paper is half of it. The other half is seeing the page the way an unmatched visit will, including the awkward states.
See the fallback before your visitors do
ReactWoo Geo keeps a default for the visit that matches nothing, and for the visit where detection fails. That behaviour is in every edition, not only the paid ones.
With Geo Core (free), important content does not have to depend on detection alone. If a country cannot be detected, or no rule matches, a sensible default stays visible. Before a rule goes live you can check setup health and preview visitor context.
Geo Core Pro is where you check both states on purpose. You simulate the context, confirm the fallback, and publish once the matched experience and the default are both deliberate. If a condition cannot be evaluated, or no rule matches, the default content remains visible.
Geo Commerce is the WooCommerce add-on, and it needs Geo Core. When a commerce rule does not match, the default catalogue remains in place. You can preview the catalogue, the product page and checkout before the rule goes live, so a visibility or pricing rule does not leave a hole where the ordinary shop should be.
The takeaway
The variant you are proud of is not the variant most people meet. Treat that as a planning rule, not as a statistic nobody has measured for your site. Write the default until it would be a respectable page with every rule switched off. Then personalise one part, and look at the fallback before you publish. A clever rule on a weak default personalises the exception and neglects the visit.