Spintax for cold email: syntax, examples and fallbacks

Spintax is a compact way to describe several possible versions of a message. In a simple flat group such as {Hi|Hello}, the braces mark the group and the vertical bar separates the alternatives. A renderer chooses one option. The surrounding text stays the same.

A greeting variation is not personalisation. It tells you nothing about whether the person is relevant or why the offer matters to them. Personalisation uses verified recipient context; variables are simply one way to insert that data. A fallback supplies a usable alternative when a field is unavailable.

Keep these layers separate in your working document. Decide what must be true about the recipient, what can vary without changing the offer, and what should happen if required information is missing. Essential context may require holding the contact out of the campaign rather than substituting generic praise.

Start with a message you would be comfortable sending unchanged to one appropriate recipient. Establish a verified reason for writing, a concrete offer and one understandable question. If the base email is vague, adding twenty openings will multiply the uncertainty.

In this illustrative example, the writer has checked a company careers page and can genuinely provide the described service. Keep that observation, the service and the sender identity fixed. The greeting and final question can vary without making a different promise.

Save the original version. It gives your team a reference when someone adds a new alternative later. Treat a changed price, a new outcome claim or a different call to action as a substantive copy change that needs its own review.

Two groups with two alternatives each create four combinations: 2 × 2 = 4. Add a third group with three alternatives and the total becomes twelve. The count describes combinations, not necessarily unique messages: repeating the same alternative can produce duplicate output.

Use the lab to see how quickly the review task grows. Its preview moves through the combinations in a fixed order, so you can inspect them deliberately. That is a teaching aid; it does not reproduce a sending platform’s random selection or predict how often each version will be delivered.

Start by reading all four examples. Then change one alternative and inspect the results again. Keep phrases self-contained: switching an entire question is usually easier to check than combining independently varied fragments of one sentence.

Similar-looking punctuation does not mean that two tools use the same template language. Instantly documents a RANDOM expression inside double braces. Smartlead documents single-brace alternatives and separate double-brace variables. Use the current instructions for the tool and campaign editor you are working in.

The examples below show those basic patterns. Field names must match the data your campaign actually maps. An example named firstName does not prove that your import has that field. Review the resulting message inside the tool after import and before launching.

The local lab above intentionally uses the simpler flat format. A parser error there does not mean a supported expression is invalid in your sending platform.

Prepare three test contacts before reviewing the whole list: a populated name, an empty value and a value containing only spaces. Separately, check that the CSV column is mapped to the field your template expects. Missing values and incorrect mapping are different problems; do not assume a fallback resolves both.

Smartlead documents conditional fallback blocks using if, else and a closing tag. The illustrative block below selects a greeting for a populated first_name field and a generic greeting otherwise. Run it against your imported test records to check the actual rendered result.

A fallback should remain true. Replacing a missing company name with “your impressive business” adds unsupported praise without fixing the missing context. If your opening depends on a verified job vacancy and that evidence is absent, stop that record for review. Do not manufacture the observation.

Once the fallback looks right, preview the complete subject and body. A correct greeting can still sit above an unresolved company token or a blank sentence later in the email.

Spintax alternatives are often chosen independently. If one group changes the subject of a sentence and another changes its verb, some combinations may become ungrammatical. Put dependent wording in the same group, or keep it fixed.

For example, “{Your team|Your account managers} {is|are} hiring” creates four outputs, including two with mismatched agreement. Keep the dependent words together: “{Your team is|Your account managers are} hiring”. Use this only when both subjects are accurate; grammar checks cannot establish whether a statement is true.

Also inspect spaces, punctuation and the transition into the next sentence. Read the message on a narrow screen, with the real signature and opt-out treatment included. A polished body in a document is not the same as the complete email your recipient sees.

Randomly changing several phrases at once makes it harder to understand why one message received a response. If you want to compare a concrete observation with a problem-led opening, create deliberate variants with a clear question behind the experiment. Keep a record of what changed and who received each version.

Decide what you will assess before sending. Relevant replies, qualified conversations and objections can tell you more about the offer than a collection of mixed response types. Use your actual reporting and campaign context; this guide does not prescribe a universal sample size or a guaranteed winning variant.

Spintax should not be used to disguise an irrelevant message or route around a recipient’s preferences. Wording changes do not repair authentication, inaccurate data or poor audience selection. Google’s sender guidance covers authentication, recipient experience and sending practices; it does not make word rotation a substitute for those foundations.

A campaign handover should include the base message, approved alternatives, required fields and evidence that the final preview was checked. Name the person responsible for changing copy after approval. Otherwise, an unreviewed phrase can quietly enter a previously checked template.

Copy the worksheet below into your campaign notes. Save a few rendered examples with the source template, including a missing-data case. When you change tools, import mappings or sequence steps, repeat the preview check rather than assuming an earlier result still applies.