When personalization tokens break: Fixing Hi {firstName}

Nothing announces this went to a thousand people faster than Hi {firstName}, at the top of an email trying to feel personal. One broken tag erases the whole illusion, and the prospect stops reading right there.

Why a broken token is fatal

Personalization tokens exist to make an email feel written for one person. When one breaks visibly, it does the exact opposite, proving in a single glance that the email is automated and untargeted.

The damage is total and immediate. Everything after Hi {firstName} is read, if at all, by someone who already knows they are on a list. The relevance you worked for is gone before the first sentence.

The common ways they break

Most failures come from data, not the tool.

• Missing data, so the raw token or a blank appears.
• Wrong-case or malformed names, hi JOHN, hi john.smith.
• Mismatched fields, the company in the name slot.
• A default that shows, Hi there or Hi [First Name] as a giveaway.

Clean data is the real fix

Tokens break because the data behind them is dirty. The durable fix is upstream: a verified, well-formatted list where names are present and correct. This is the same discipline as verification and list building, applied to the fields you personalize on, not just the address.

Use fallbacks that pass as human

Set a default for every token that reads naturally when the data is missing. If a first name is absent, the fallback should be a normal greeting, not Hi {firstName} or Hi there in a way that flags the gap. A good fallback is invisible; the reader cannot tell the token did not fire.

Better still, do not personalize on a field you cannot rely on. A token you are unsure of is a risk; a clean generic line is safer than a broken personal one.

Normalise the data before it reaches the sending tool

Fallbacks catch missing data. Malformed data is worse, because the token fires and still looks wrong. Run every list through a normalisation pass before upload:

• Case: convert JOHN and john to John. Watch names where simple title-casing fails, like McDonald or van der Berg.
• Stray content in the name field: strip titles, credentials and emojis. "Dr John Smith MBA" should merge as John.
• Split fields correctly: a full name pasted into the first-name column produces Hi John Smith, which reads almost as badly as a raw tag.
• Company names: shorten legal forms for the body copy. "Acme Logistics Sp. z o.o." reads like a database export; "Acme Logistics" reads like a human wrote it.

This pass belongs in the same routine as list maintenance: data that was clean at import decays as roles change and records get merged.

The European complication: diacritics and encoding

Campaigns across Poland, Germany, Lithuania and the Nordics add a failure mode most US-centred advice ignores: character encoding. A Paweł that arrives as Pawe? or PaweÅ‚, a Jürgen rendered as Jurgen, an Ąžuolas mangled beyond recognition, each one signals sloppy mass mail to the reader even though the token technically fired. Export and import lists as UTF-8 end to end, and spot-check names with diacritics in the test send. Ripe Leads runs multilingual campaigns in LT, EN, DE and RU, and treats diacritic checks as a standard QA step, because a mangled name in a German or Polish inbox costs more trust than no name at all.

Custom tokens raise the stakes

First name and company are the low tier. Real gains come from custom tokens: the role a company is hiring for, a trigger event, an opening line researched per prospect. These make the email land harder, and they break harder too, because a wrong custom line is worse than a missing one. Congratulating a company on an expansion that was a layoff is not a formatting error, it is a credibility error. The rule: any custom field gets a human skim before sending, and any row where the researched line is doubtful gets the generic variant. How to build these fields without drowning in research time is covered in personalization at scale and the prospect research workflow.

If you are thinking nobody notices

Here is the reality: your prospect receives cold email daily and has seen every broken tag pattern there is. Decision-makers pattern-match in under a second, and Hi there plus a vague compliment is a known signature of automation. Reply rates on cold email typically sit in the low single digits; a visibly broken campaign lands at the bottom of that range and adds spam complaints on top, which turns a copy problem into a deliverability problem. The two minutes of testing are cheap against that.

Always test before you send

The last line of defence is a test send to yourself and a couple of colleagues before every campaign, checking that tokens resolve and fallbacks read naturally. It takes two minutes and catches the mistake that would otherwise reach hundreds of prospects. Skipping it is how Hi {firstName} goes out at scale.

Make it a fixed pre-flight routine: send test rows with full data, missing data and diacritics; open on desktop and phone; read every merged line aloud; then launch. The same discipline that prevents the other reply-killing mistakes prevents this one.