The product title playbook
The title is the highest-leverage line of text on every product: it decides search matching, click-through, and on-page confirmation at once. Most catalogues fill it with internal jargon.
Product titles fail in predictable ways: named for internal SKUs, keyword-stuffed into unreadability, or duplicated across variants until nothing distinguishes anything. This playbook fixes titles as a system — derive the vocabulary from how buyers actually search, apply a per-surface structure (your store, Google results, marketplace listings have different rules), and roll it out catalogue-wide with the measurement to prove it helped.
Stage 1: Derive the vocabulary
The words come from buyers, not the brand. Mine your own search-terms data, the queries that already land on each product, and the language in reviews — customers describing the product in their own words are writing your title for you. Note where buyer vocabulary and your catalogue vocabulary diverge; that divergence is invisible lost traffic.
On-site search terms and landing-page queries per product
Review language — how owners actually name the thing and its qualities
The attribute order buyers specify: brand, type, spec, size, colour — category-dependent
Divergences between internal naming and buyer naming, listed
Stage 2: Structure per surface
One title does not serve every surface. The store title leads with what confirms the click and differentiates within your catalogue. The SEO title works within pixel limits and leads with the search-matched phrase. Marketplace titles follow that marketplace’s length and structure rules, where attribute completeness gates filter visibility. Same facts, per-surface arrangement — from one source of truth so they never contradict.
Stage 3: Roll out with measurement
Catalogue-wide title changes deserve care: ship in batches, watch impressions and click-through per batch in Search Console, and keep the old titles recorded for rollback. Expect impression shifts within weeks. Where a batch loses, the record shows exactly what changed — which is what makes batch rollout safe where a big-bang rewrite is not.
How it works
Mine the language
Search terms, landing queries, and review vocabulary, per product and category.
Apply per-surface structure
Store, SEO, and marketplace variants from one fact base, to each surface’s rules.
Batch, measure, keep receipts
Rolled out in batches with per-batch measurement and recorded rollback state.
What you get
Titles in the language buyers actually type
Per-surface variants that stop leaving marketplace filter traffic on the table
A rollout that can prove which changes helped
Internal jargon retired from customer-facing text
Handled by these agents
Works with
Frequently asked questions
How long should a product title be?
Per surface: SEO titles within display limits (roughly 50–60 characters before truncation), marketplace titles to that marketplace’s rules, store titles as long as clarity requires and no longer. Length is a constraint, not a goal.
Should titles include the brand name?
When buyers search with it or it differentiates, yes — usually leading on marketplaces, often trailing on your own store where the context is implied. The search-terms data answers it per category.
Do title changes risk existing rankings?
A title aligned with actual query language nearly always helps, but batch rollout with recorded rollback exists precisely because "nearly always" is not "always". Measure per batch.
Which agents run this?
Visibility derives the vocabulary and owns the measurement; Builder ships the changes across store and marketplace surfaces from the shared fact base.
Keep reading
Product page optimisation
Rebuild the pages that decide whether a visit becomes an order, ranked by revenue at stake.
Marketplace listing optimisation
Keep listings competitive on marketplace search, consistent with your store, and honest to spec.
Write collection copy
Collection page copy that helps buyers choose — and gives the page a reason to rank.
Plan a BFCM campaign
The peak-sales campaign programme: offer architecture, list warm-up, sequencing, and load readiness.
Put the team to work on your store
Connect your storefront, analytics, and email stack, set a goal, and let the agents run the work end to end. Start on the free plan with your own model key.