Your Ecommerce Growth Strategy Has a Truth Problem Before It Has a Traffic Problem

An ecommerce business should not increase paid or organic acquisition until one important product can travel from advert or listing to checkout without unexpectedly changing identity, price, availability or fulfilment terms.

That is my line in the sand. Traffic does not repair a confused offer. It increases the number of people exposed to it.

Growth marketing teams often inherit the cleanest part of the story. Google Ads reports the click. SEO reports the impression. Social media reports the engagement. The ecommerce platform reports the basket and order. Somewhere between those systems, the customer sees a product that changes beneath their feet, and every dashboard can still look locally reasonable.

I call this offer continuity: the ability of a commercial promise to remain recognisably the same across every surface required to complete it. The product does not need identical wording everywhere. It does need stable identity, price logic, availability and material conditions. Before discussing more customer acquisition, I would test offer continuity on one commercially important variant. Not a category average. Not a feed-health percentage. One item, followed all the way through.

The category average is where broken variants go to hide

A category can look healthy while the promoted size, colour or configuration is unavailable. A feed can contain the correct parent product while its link loads the wrong variant. A product page can display an attractive sale price until the customer selects the option shown in the advert. This is why I distrust reassuring averages at the start of an investigation. Averages describe a population; customers attempt to buy a specific thing. Select the exact variant being promoted or discovered and record:

• the store’s product and variant identifiers;
• the submitted title and destination URL;
• target country, language and currency;
• base price, sale price and active promotional dates;
• availability and fulfilment restrictions;
• the variant selected when the page first loads;
• the delivery or collection promise shown before purchase;
• what enters the basket;
• what reaches checkout;
• date, time, device and relevant account state.

Do not quietly switch to a working variant when the advertised one fails. That instinct, to rescue the test, is exactly how operational truth gets replaced by presentation truth. Google Merchant Center requires submitted landing pages to show the relevant product clearly and align key details such as product, price, currency and availability with the submitted data. When variants appear on the same page, the submitted product should remain the primary focus (Google Merchant Center: landing-page requirements). That is a Google Shopping requirement, not a complete conversion rate optimisation strategy. But it supplies a useful minimum question: did the click arrive at the thing that was actually offered?

Follow the Offer Continuity Route

The fastest useful audit is a route, not a store-wide crawl. Follow one variant through five surfaces and stop at the first material disagreement.

If the product source is unresolved, a feed rule cannot rescue it. If the source is correct but the export is stale, rewriting the page is not the first move. If the page and feed agree but checkout changes the price or rejects fulfilment, buying more traffic is an expensive way to rediscover an operational limit. This distinction matters because “price mismatch” is an observation, not a cause. The cause may be a promotion schedule, tax rule, currency setting, variant map, cache, inventory connector, theme, app or manual edit. I would record where the promise first changes before letting a team name the mechanism.

Price is part of the identity of the offer

Marketers often treat price as a field attached to a product. Customers experience it as part of the product they were invited to consider. If the active price becomes a scavenger hunt between a feed, crossed-out figure, promotional banner, basket rule and surprise checkout charge, the commercial identity has fractured. The problem is larger than a data warning. The customer no longer knows which offer is real.

For UK businesses, price continuity also deserves appropriate legal attention. The Competition and Markets Authority’s price-transparency guidance covers total prices, mandatory fees and charges, drip pricing and partitioned pricing. The guidance was published on 18 November 2025 and updated on 7 January 2026 (CMA: price transparency guidance). The precise legal position depends on the offer and circumstances, so a marketing or technical audit should flag material discrepancies for qualified review rather than declare compliance from a crawl. Run the commercial check in sequence:

1. Record the approved base and sale prices for the exact variant.
2. Record the values submitted to the relevant product platform.
3. Open the submitted landing URL under the target-market conditions.
4. Confirm the loaded variant and visually dominant price.
5. Add that exact item to the basket.
6. Continue through checkout far enough to expose mandatory charges and available fulfilment, without placing an order unless the test authorises it.
7. Preserve the evidence with its date, device, market and account state.

Do not staple an invented lost-revenue figure to the finding. A discrepancy can justify correction without a fictional commercial impact model. If the business wants a loss estimate, that becomes a separate measurement problem with its own evidence.

Availability means the transaction can be honoured

“In stock” should mean more than a green label generated by whichever system refreshed last. Google’s landing-page guidance says the page should clearly show availability for online purchase. If a product can be ordered, an active buying control should exist; preorder or backorder offers should show an expected dispatch date; collection-only or other delivery restrictions should be disclosed (Google Merchant Center: landing-page requirements). That exposes several continuity failures: a submitted item marked available whose target variant cannot be selected; a product page that accepts the item while checkout rejects the customer’s location; a category badge saying available while the preselected size is sold out; a backorder with no expected date; inventory changing faster than product data refreshes; a collection-only restriction appearing after commitment.

An audit should not invent the retailer’s stock policy. It should identify what each system states, which source is authorised and whether the customer-facing route matches what the business can fulfil. Fast-changing inventory creates an important boundary. One captured mismatch proves that the recorded surfaces disagreed at that moment. It does not prove duration, frequency or cause. Repeat observations may be needed before a business changes an entire growth strategy.

Checkout is a marketing surface whether the marketing team owns it or not

The organisational chart may place checkout outside marketing. The customer journey does not care. If the selected item changes, a promotion disappears, delivery becomes impossible or a mandatory charge arrives late, the original acquisition promise has failed before conversion. The implementation owner may sit in ecommerce operations, development or finance, but the evidence belongs in the decision about whether more demand should be released. Google’s product-data guidance recommends mapping the customer journey from query to checkout and maintaining accurate price and availability across product data and landing pages (Google Merchant Center: product-data optimisation). That platform advice points to a broader branding truth: a brand promise is not the sentence in the advert. It is the experience the organisation can carry to completion.

This is where website design becomes commercial infrastructure. Variant selection, price hierarchy, stock language, delivery disclosure and error recovery are not decorative details. They govern whether the customer can understand and complete the offer. I would therefore include checkout evidence in a performance marketing review even when the marketing team cannot implement the fix. Accountability and relevance are different questions. Marketing may not own the defect, but it should not pay to amplify it blindly.

Diagnose the first break before choosing the fix

Product truth crosses teams, which makes blame an attractive shortcut. Resist it. Use the first-disagreement rule: if the approved identity, price, promotion or fulfilment rule is unsettled, merchandising or trading must settle the offer. If the approved source is clear but submitted data differs, inspect the export, mapping and update schedule. If the submission agrees but the product page changes the state, inspect ecommerce platform and website behaviour. If the page agrees but the basket or checkout changes it, investigate application, promotion, tax and fulfilment logic. If every displayed value agrees but the order cannot be fulfilled, the issue has reached inventory, location or delivery operations.

This sequence prevents three teams fixing three symptoms while the first broken handoff remains in place. It also stops the technology stack becoming the automatic villain. A WordPress or Shopify store can be configured well or badly; an app can expose a conflict without being its origin. Name the first changed value and the responsible process before prescribing a platform migration, redesign or plugin removal.

Put acquisition behind a continuity gate

Once the first disagreement is understood, make the traffic decision at the same scale as the evidence. Pause expansion to the affected variant when the route cannot honour the promoted identity, price, availability or material conditions. That does not automatically justify shutting down every Google Ads campaign, organic listing or social promotion for the product family. When the exact variant passes the route under recorded conditions, release a limited test. Define the audience, spend or traffic boundary, the conversion event and the condition that triggers another continuity check. I use this gate:

Passing the gate does not guarantee conversion, approval, visibility or profitable business growth. It makes the performance test interpretable. That is enough. Growth strategy improves when each stage earns the right to introduce more volume.

Conversion optimisation starts by preserving the promise

Conversion rate optimisation is often reduced to button colours, persuasive copy and shorter forms. Those interventions can matter, but they arrive too late if the offer itself changes during the journey. My priority order is:

The order is not universal. A stable mature store may already have the first three under control and gain more from interface experimentation. But where offer continuity is uncertain, persuasion tests can optimise the response to a moving target. This is the point where branding, ecommerce operations and digital growth meet. A logo system may create recognition. Campaign creative may create desire. The purchase route decides whether the organisation behaves like the same brand when money is involved. I would not approve a growth plan that celebrates customer acquisition while treating that handoff as somebody else’s metric.

Keep the route replayable

A one-off pass is useful; a replayable route becomes an operating asset. For each important variant, retain the approved source values and accountable owner; submitted values and source timestamp; exact destination URL and loaded variant; product-page, basket and checkout observations; market, currency, device, login and consent state; the first disagreement and where it appeared; correction and deployment time; repeat-test result; acquisition decision; and the condition that will trigger another check.

Recheck when promotions begin or end, stock systems change, feeds migrate, checkout applications are replaced, tax or shipping rules change, a website redesign launches or a new market is opened. Do not rerun the whole estate by ritual. Replay the routes most exposed to the change. The honest limitation is that agreement across the checked route does not prove every customer will see the same state. Location, device, consent, account, currency and timing can matter. That is why the record contains conditions and why a limited release is more intelligent than a victory lap.

Before asking what another thousand visits might deliver, take one important product from its authorised source to checkout. If the offer changes, fix the first broken handoff. If it holds, release a measured test. Customer acquisition should amplify a coherent commercial promise. It should never finance the search for one.