Social marketing for PrestaShop: channels, ads and measurement

What we would do

We would choose a customer problem and a channel your team can maintain, then prove that the shop can fulfil the promise and measure the resulting orders before expanding. The route depends on your existing PrestaShop version, theme, checkout, installed modules, consent platform, catalogue data and ERP or marketplace integrations, as well as your markets, margins and time for replies. Keep a working integration when it meets the requirement; a different module will not resolve a missing consent choice or an unclear purchase-status policy by itself.

For changes to tracking, feeds or storefront displays, work on a private staging copy with the same relevant configuration. Isolate it from live posting, production advertising data, customer messages and operational exports. Rehearse the consent, product-matching and order tests with supported test tools, record how to reverse the change, then make the planned switch with a named person checking live results. Publishing an ordinary post does not require a shop rebuild.

Choose the connection you actually need

For discrepancies involving Merchant Center, GA4 and Google Ads, the PrestaShop Product Feed & Tracking Audit documents feed and measurement faults; corrections and campaign management are excluded.

Choose your channels and define a test

Start with a channel you can support with useful content, customer replies and measurement. Add another when you have capacity. A regular schedule helps you organise the work and compare results; it does not guarantee reach.

Use these as content ideas. Check whether your intended buyers use the channel in your market. A price, age bracket or B2B label alone cannot determine the result. A premium product can be demonstrated in short video; a consumer brand can have a relevant professional audience on LinkedIn.

Plan a manageable experiment

1. Choose the product range, market and customer problem you want to address.
2. Choose a goal: qualified inquiries, website purchases, repeat orders or usable content assets.
3. Set a maximum cash budget and record the time you can spend on production and replies.
4. Prepare the landing pages, tagged links and reporting before publishing.
5. Test useful explanations, demonstrations, customer questions and product offers. Keep a record of what changed.
6. Choose a review period that covers your buying cycle. Include conversion delays and returns when judging results.

You can schedule image content in batches or turn one demonstration into several formats. Allow time for moderation and maintenance on every channel. Avoid universal posting quotas, fixed conversion-rate cutoffs or a mandatory number of videos. A few orders can change a small sample sharply.

Example: a kitchen-storage range with limited staff time

If customers ask how to organise a small kitchen and your team already has useful room photographs, test Pinterest links to a relevant selection or guide. If the main uncertainty is whether a fitting works, demonstrate it on video and link to the exact product with its dimensions and installation requirements. Choose from the question buyers need answered and the material you can produce, then assess the tagged visits and orders against the agreed goal.

Write a short test brief: the buyer and market, the product selection, the promised answer, the destination, who replies to questions, the cost limit and the condition for continuing. Keep the offer, stock and delivery terms current while the test runs. If you change the creative and the destination together, record both; you will have less evidence about which change explains the result.

Set up conversion tracking and consent

Tagged links support acquisition reporting in an analytics implementation. Advertising pixels and server APIs send events to their respective platforms for measurement and eligible advertising uses. Install the integrations you need for a defined purpose; you do not need every platform's pixel simply because you run a store.

Check the events that match your customer journey

• Product views should identify the product or combination shown.
• Add-to-cart and checkout events should represent actual actions, with the right quantities and values.
• Purchase events should use the actual order, currency, value and items. Define the order status that qualifies as a purchase for your reporting.
• Catalogue item identifiers and event item identifiers must agree for product matching.

An order-confirmation page is not a universal proof that payment has completed. Bank transfer and other delayed payment methods need an explicit reporting policy. Check that refreshing a confirmation page, returning from a payment provider or retrying a server request cannot create extra purchases.

Deduplicate browser and server copies

When you send the same action through a browser pixel and a server API, use a stable event identifier for that action and match the event name. For Meta, the browser parameter is eventID and the server parameter is event_id. TikTok uses event_id for this purpose. Check receipt and deduplication in the platform's event tools.

Use a different identifier for a different action. Browser and server delivery can complement each other, but server delivery does not guarantee complete measurement. For module-specific configuration, use the Meta configuration help or the TikTok configuration help.

Verify consent on both delivery paths

Under applicable European ePrivacy and cookie rules, advertising trackers generally need prior consent. Personal-data processing also has GDPR requirements. An analytics-only choice does not authorise advertising tracking. Explain the actual purposes and recipients, provide a real refusal choice and let visitors withdraw consent easily.

1. Open a fresh browser session without a saved consent choice. Check that advertising trackers are blocked.
2. Refuse advertising consent and test a product view, cart action and order.
3. Accept the relevant advertising purposes and check the intended events.
4. Withdraw consent and repeat the actions.
5. Check server requests and queued deliveries as well as browser requests.

Use a compatible consent management platform and test its integration after changes. A consent setting in a module does not remove your responsibility for configuration, disclosures or lawful processing. Moving tracking to a server does not bypass those responsibilities. Browser restrictions also differ: Chrome retained user choice rather than completing a universal third-party-cookie removal, while other browser policies and Apple's app-tracking rules are separate.

Rehearse the integration before the live switch

Start by listing every source of the same tracking tag: standalone modules, a tracking suite, the theme, a tag manager and any custom checkout code. Give each destination an intended owner. A module's duplicate-tag detection can only cover the integrations it recognises; inspect actual browser requests and server deliveries rather than assuming an old snippet disappeared.

1. Record the current module releases, shop context, account identifiers, consent mappings and purchase-value policy. Keep credentials out of screenshots and shared reports.
2. On the private copy, disable automatic social publishing and isolate catalogue schedules, analytics destinations, emails and order exports. Confirm the copy cannot overwrite a production feed or report rehearsal orders as real sales.
3. Use the integration's supported testing mode or a separate test destination. Check the live theme and checkout journey, including combinations, promotions, currency, language and delayed payment methods that your customers use.
4. Run the refusal, acceptance and withdrawal checks above. Follow an eligible order from the shop record into the platform's event tools, comparing the event identity, items, amount and currency. Repeat the confirmation-page visit and supported delivery retries.
5. Record the expected and observed result for each case. Investigate a missing or extra event before altering campaign bids to compensate for it.
6. Switch during an agreed window, removing the superseded sender as the replacement takes ownership. Check credentials, shop identifiers and schedules again on production, then verify the agreed live journey and monitor incoming orders.
7. If checks fail, pause the affected change or campaign and follow the recorded reversal steps. Recheck tracking ownership after restoring settings, so rollback does not leave competing senders active.

What to ask the developer or agency

• Which component owns each event, feed and publishing action, and which existing snippets or modules overlap?
• What qualifies an order for purchase reporting, which value components are included, and how are cancellations or refunds reflected?
• Where is evidence for consent refusal, withdrawal, browser/server deduplication and repeated payment returns?
• Which product identifier connects the ordered combination, the event and the destination catalogue?
• Who controls the business account, access recovery and connection renewal after handover?
• Where are failed deliveries or feed refreshes visible, who responds, and what are the switch and reversal steps?

Tag campaign links and read attribution correctly

Use a stable naming convention for social links. Keep source, medium and campaign values consistent, and use content to distinguish a creative, creator or Pin. For example:

?utm_source=pinterest&utm_medium=social&utm_campaign=spring_collection&utm_content=pin_bathroom_inspo
?utm_source=instagram&utm_medium=influencer&utm_campaign=spring_launch&utm_content=anna
?utm_source=tiktok&utm_medium=social&utm_campaign=spring_demo&utm_content=video_042

Append the parameters to the public product, category or content URL. Use & when the URL already has a query string. Avoid session or authentication tokens. Never put customer names, email addresses or other personal information in campaign parameters.

Without UTMs, analytics may still recognise a social source through the referrer. Apps, redirects and privacy controls can limit that information. Tags give you deliberate campaign detail; they do not guarantee that every visit or later order can be attributed.

Keep canonical metadata separate from the landing URL

A product page can use a clean, untagged rel="canonical" while the browser's landing URL contains UTMs. The canonical tag tells search engines which URL you prefer. It does not itself remove parameters from the address bar or stop analytics reading them.

If campaign detail disappears, follow the actual click through shorteners, redirects, language selection and URL rewrites. Check that the landing parameters remain available until your analytics implementation captures them. Test a real published link on a phone as well as a desktop.

Use the right report and scope

In a configured GA4 implementation, use Traffic acquisition with session source, medium and campaign to investigate sessions. Event-scoped attribution in advertising reports uses the property's attribution model, which defaults to data-driven attribution. Those reports answer different questions. A custom medium such as influencer also needs a deliberate reporting rule if you want a dedicated channel group.

Configure ecommerce purchase reporting and a stable transaction_id before expecting purchase revenue alongside acquisition dimensions. UTMs alone do not install analytics, create ecommerce events or add campaign reports to native PrestaShop statistics. See PrestaShop analytics for the broader reporting setup.

Native visitor-origin statistics depend on the installed statistics modules. Legacy Referrers configuration, where available under Shop Parameters > Traffic & SEO, uses referrer and request-URL patterns. Keep it separate from GA4 campaign acquisition and purchase attribution, and check availability on your release.

Prepare product catalogues for social platforms

A catalogue export supplies product data to a platform. An embedded social feed displays posts on your storefront. A tracking pixel reports visitor actions. These are separate connections, even when one integration offers more than one of them.

Build exports from your maintained PrestaShop catalogue. Reuse the underlying data across channels, but validate each destination's required fields, formats and policies. A Google Shopping export is a useful starting point, not an automatic guarantee of Meta or Pinterest compatibility.

For dynamic product ads, check both catalogue diagnostics and event-to-item matching. Keep out-of-stock items correctly marked rather than presenting them as available purchases. Stale feeds, failed refreshes and mismatched variants can leave availability or prices wrong.

When using Meta product discovery features, verify business and market eligibility and the website checkout destination. Do not plan around an assumption that a Facebook or Instagram Shop provides native checkout. See PrestaShop merchant feeds for export design and validation.

Trace a combination through the whole chain

Take a storage box offered in different sizes. Open the advertised size on a phone, note its selected combination, price, currency and availability, then find that same offer in the export and the destination's ingested catalogue. Trigger the permitted product and cart events and compare their item identifiers. If the catalogue contains size-specific offers but the event sends only the parent product ID, the platform may not match the intended item.

After a stock or promotional-price change, check both the generated export and the last ingestion result. A correct file does not prove the platform fetched it successfully. Assign someone to review failed refreshes and rejected products; otherwise a working launch can become a stale catalogue while the ads continue spending.

Start Facebook and Instagram ads

Confirm that your selected conversion source reports the actions you intend to optimise before spending. Meta browser events include ViewContent, AddToCart, InitiateCheckout and Purchase. Check actual values and product matching, and check deduplication if you also send Conversions API events.

Match the objective to the outcome

Use a Sales objective when your goal is purchases, with the appropriate destination and conversion event. Prior pixel history is not a fixed prerequisite of a particular number of purchases. Traffic optimisation targets visits rather than proving purchase demand; use it for a deliberate traffic goal. Engagement campaigns serve a different goal again.

Choose prospecting audiences deliberately. You can test relevant interests, eligible lookalikes or broader targeting within your serviceable market. For example, a yoga-mat shop might test fitness interests against a broader audience. Do not assume a particular audience size, a 1% lookalike or broad targeting must win.

Test different creative explanations

• Compare a clear product image with a product-in-use image.
• Try a demonstration that shows the product's purpose early. Add captions so the message works without sound.
• Use a carousel for different products, product angles or a problem-and-solution sequence. Check each card's destination.
• Use catalogue-based ads when your catalogue and event identifiers are ready.
• Adapt assets to the placement. Check cropping, readable text and the mobile landing experience.

Set URL parameters on the relevant ads

Configure the URL parameters for each relevant ad or creative, and check them after publishing. Do not assume one campaign setting tags every ad in an account. An example parameter string is:

utm_source=facebook&utm_medium=paid_social&utm_campaign=spring_launch&utm_content=yoga_mat_demo&utm_term=fitness_test

Use your own labels, or Meta's dynamic URL parameters such as utm_campaign={{campaign.name}}, utm_content={{ad.name}} and utm_term={{adset.name}}. Set them on each relevant ad and check the populated values after a published click; they do not configure the rest of the account. Maintain a stable naming convention and test the published destination. If you need reporting by placement platform, configure and verify that tagging choice rather than labelling every placement Instagram or Facebook indiscriminately.

Budget for a controlled test

Set an affordable loss limit using your expected acquisition cost and contribution margin. Avoid splitting a small budget across many nearly identical ad sets. Campaign-level budget allocation can distribute spend across ad sets; ad-set budgets give you a different level of control.

Allow for reporting delay and your buying cycle before judging results. Significant edits can disturb delivery and learning, so record changes and avoid making several at once. There is no universal budget-increase percentage that guarantees stable delivery, and a fixed daily amount cannot guarantee enough purchase data.

Build and evaluate retargeting audiences

Retargeting reaches eligible people who previously interacted with your store or content. Website audiences use the platform's permitted website signals. Other audience sources can include platform engagement or properly collected customer lists, subject to the platform's rules and your permissions.

A website pixel does not observe everyone who sees your social content. An audience created in one advertising platform does not automatically become an audience in another.

Control repetition and duration

• Choose audience membership windows from your buying cycle and test them. A low-cost consumable and a furniture purchase may need different treatment.
• Use frequency caps where the campaign type provides them. Otherwise manage spend, audience windows, exclusions and creative rotation.
• Exclude recent purchasers from campaigns that continue selling the same item. Treat cross-selling separately.
• Try a sequence: product reminder, useful explanation or permitted review, then an incentive if the margin supports it.
• Watch frequency together with reach, click quality and purchases. A single frequency number does not diagnose fatigue in every campaign.

Fix delivery surprises, payment failures or checkout usability directly. Ads cannot repair those problems. See PrestaShop checkout for the store-side work. Abandoned-cart email is a separate channel with its own eligibility and consent considerations.

Measure additional sales rather than assuming them

Visitors who already know your store may return without an ad. A low reported cost per purchase does not show how many sales the ads added. Compare audiences, discounts and contribution margin. Where practical, use a properly designed holdout test that leaves an eligible comparison group unexposed.

Read click-through and view-through attribution separately. An impression followed by a purchase is an attribution signal, not proof that the impression caused the purchase. Do not assume retargeting, cart audiences or any one platform is always the cheapest acquisition route.

Use Pinterest for product discovery

Use Pins to connect visual ideas with relevant products or explanations. A home, fashion, food, beauty, wedding or gift catalogue gives you possible themes to test. Technical or B2B products can also work when you can show a useful use case. Check actual search interest and results rather than excluding a whole category.

Prepare metadata and a catalogue

Product Rich Pins can draw current product details from supported page metadata. Check what your live theme outputs, including product title, price and availability. Add missing metadata through your theme or an appropriate module. Pinterest can enrich eligible Pins automatically; inspect the resulting Pin rather than relying on an assumed public Rich Pin validator.

A retail catalogue is a separate data source. Pinterest supports formats including CSV, TSV and supported XML. Required fields include id, title, description, link, image_link, price and availability, with item_group_id for products with variants. Supply other identifiers where applicable. Check business-account, website-claiming and merchant eligibility requirements.

Use the catalogue's data-source validation and diagnostics. Keep a hosted export available for scheduled ingestion and inspect refresh errors. Reusing your product data saves maintenance, but check the Pinterest mapping before reusing an export from another destination.

Build useful Pins

1. Choose the specific product, category or helpful content page that answers the Pin's promise.
2. Add and test a destination link. A link is an option, not an inherent feature of every Pin.
3. Use a suitable vertical image. A 2:3 image such as 1000 by 1500 pixels is a practical starting format.
4. Write a clear title and description using relevant product, material, colour, room and use-case terms.
5. Test different views and explanations of the same product without making misleading claims.
6. Add campaign tags and check the published mobile destination.

For example, describe a lamp as a modern brass desk lamp with an adjustable arm and warm LED light only if those details are true. Compare a clean catalogue image with a real-room image instead of assuming either always performs better.

Read Pinterest outbound clicks alongside tagged sessions and configured ecommerce purchases. Saves and impressions help diagnose discovery, but do not equal orders. Choose a review period that allows for your customers' planning and buying cycle; searchable content has no guaranteed lifespan or ranking.

If you also want to show an inspiration board inside the shop, Pinterest Feed Integration displays selected Pins and boards, with clicks opening the original Pin. Treat that storefront placement as a separate test: check whether it helps the buying task and whether it draws visitors away from the intended product journey.

Measure TikTok website and Shop sales

Separate the route to purchase before comparing revenue. TikTok can report attributed website purchases when your event setup supplies the purchase and value. A TikTok video-feed display on your store does a different job from Pixel or Events API tracking.

Shop orders are not automatically native PrestaShop orders without an integration. A suitable connector can import them. Verify stock, order-status, refund and customer-data mapping before relying on the synchronisation. Marketplace CRM features and exportable contact data are different: eligible US TikTok Shop sellers have platform messaging options including email, but that does not grant an unrestricted mailing list for your own use.

Use the current purchase event correctly

Use Purchase for a new TikTok setup. TikTok renamed CompletePayment to Purchase in May 2025, and existing legacy event names remain supported. Keep the event names consistent across browser and server delivery, and send a shared event_id when both report the same purchase.

Check the real order value, currency and item identifiers in Events Manager. Define when an order qualifies, and test duplicate delivery and payment returns. A loaded pixel without the intended purchase event cannot supply the purchase reporting you expect.

Keep audience sources separate

Use TikTok engagement audiences for eligible video viewers and other on-platform interactions. Use website audiences for tracked website visitors. Someone who only watched a video has not generated a website visit.

A TikTok pixel audience does not transfer into Meta. To retarget eligible website visitors through Meta, you need Meta's own permitted event signals or another eligible Meta audience source. Do not treat a TikTok video view as a Meta website event.

Run a measurable content test

• Show the product solving a real problem. Test demonstrations, honest before-and-after comparisons, customer reactions or unboxings where appropriate.
• Use tagged links wherever your account and content format support clickable destinations.
• Use a dedicated campaign code if the offer fits your margin, and monitor whether it spreads to other channels.
• Offer an optional post-purchase question about discovery. Treat answers as self-reported information.
• Record production costs and measure profit from resulting orders over your buying cycle. Use those results to assess profitability alongside reach, video output and audience demographics.

Combine signals by order identifier. A buyer who clicked a tagged link, used a code and named TikTok in a survey still placed one order. Keep survey-only, code-only and overlapping orders visible instead of adding the report totals.

Example: a demonstration followed by a later search

A customer sees your storage-box demonstration, later searches for your shop and uses the video's code at checkout. Record the qualifying order as a code redemption and, if supplied, a self-reported discovery source. The final measured session can still be organic search. Keep those observations together without rewriting the session source or claiming that the code proves an additional sale.

If the same customer orders inside TikTok Shop instead, reconcile that route's settlement, returns and connector-created order separately. Check whether any imported order also triggers your website purchase sender: a marketplace sale must not accidentally appear as a website checkout event because an import changed an order status.

Use LinkedIn and prepare your B2B store

Test LinkedIn when your intended buyers, specifiers or professional communities are active there. A considered purchase, supplier relationship or technical problem can give you useful content. Judge the time spent against qualified inquiries, contribution margin and the sales cycle. Order value alone is not a screening threshold, and B2C is not an automatic exclusion.

Configure the business buying experience

In Shop Parameters > Customer Settings, enable B2B mode when it fits your store. The configuration is PS_B2B_ENABLE. In core PrestaShop 1.7.8 and 9, it adds company and identification-number fields to the customer registration form. SIRET is a business identification number, not a VAT number.

The back-office customer form has additional B2B fields, including APE and credit-related settings. VAT number belongs to address data. These are different forms and purposes. B2B mode does not itself grant VAT exemption, create negotiated prices or implement a complete credit-payment workflow.

Manage business customer groups under Shop Parameters > Customer Settings > Groups. Use group reductions or specific prices for appropriate pricing, and configure category group access where needed. For basic price hiding, set Show prices appropriately for Visitor and Guest groups and allow the intended authenticated groups to see prices.

Group price display can show prices with or without tax; it does not assign a separate product tax-rule group or validate a VAT exemption. Configure tax rules, customer addresses and any VAT-validation integration separately. Quote approval and more elaborate access policies are additional requirements. See PrestaShop B2B and prices and VAT.

Publish useful material and track the pipeline

• Explain a buyer problem and the practical tradeoffs between solutions.
• Share genuine, anonymised findings from your business where you have suitable data. Check calculations and privacy before publishing.
• Use customer stories only with permission and accurate outcomes.
• Show relevant operations such as quality checks, fulfilment or technical support.
• Keep the company page accurate and use named profiles where personal expertise helps. Compare results rather than assuming one account type always gets more reach.

For outreach, research the recipient, explain why the message is relevant and respect their response. Set a sustainable content and reply schedule. No universal reply rate or minimum posting frequency applies.

Tag LinkedIn links with a consistent source and medium. If you need a discovery-source field on registration or order confirmation, add it through an appropriate extension; B2B mode does not provide that field automatically. Use the LinkedIn tracking help page for module-specific advertising measurement.

Track inquiries, qualified accounts, opportunities, orders and revenue in a CRM or spreadsheet. Keep the customer or order reference so you can match later purchases without counting the same deal twice. A long sales cycle may need a longer review than a direct-to-consumer promotion. For example, a procurement contact may read a technical LinkedIn post, request a quote through your site and later order through an approved trade account. Keep the quote, company account and eventual order linked, and record who follows up. A qualified request can be an interim result while the sale is pending; count its realised revenue only when the order meets your agreed policy.

Choose creators and agree the collaboration

Choose creators by audience relevance, content quality and a credible relationship with their viewers. Assess likely sales from audience fit and a measured test. Ask for relevant audience information, examine recent content and comments, and check that the creator can accurately demonstrate your product. Build a shortlist from customers who already mention your products, specialist communities and accounts answering your buyers' questions. Read enough recent work to distinguish useful discussion from generic engagement, and confirm that the intended audience is in a market you can serve.

Possible uses include a new product launch, entry into another language or market, a visual demonstration, or a specialist community. Define the job before selecting the account. A local specialist can be more relevant to your store than a large general audience, but test the outcome.

Ask for a quote suited to the deliverable and rights. There is no universal creator fee, gifting price ceiling or commission model. Native cart rules can supply discount codes, but they do not manage a complete affiliate programme or automatic commission settlement. In a first approach, name the product, explain why it fits the creator's work and state whether you are offering an unsolicited gift or proposing agreed deliverables. Settle expectations before dispatch; a product sent without a posting agreement is a cost you may never recover.

Use a clear written agreement

1. State the product and customer problem, and agree honest claims the creator can support.
2. Agree the deliverables, timing, payment or gift, and what happens if a deliverable is not produced.
3. Provide a specific landing page, tagged link and code where appropriate.
4. Require prominent, unambiguous advertising disclosure in the audience's language, including for relevant gifted collaborations.
5. Agree rights for the store, email, social posts and paid ads separately, including duration and any limits.
6. Define how you will review results and handle returns before commissions are settled.

Do not rely on #gifted alone as advertising disclosure. Use clear advertising wording and the platform's disclosure tools where appropriate, and check national guidance. Each post, Reel or Story needs a clear disclosure. For commercial links that the publisher controls, request rel="sponsored"; nofollow is also acceptable. Treat those links as referral traffic, without a promise of search-ranking credit.

Reuse content with permission

Licensed creator photos can join the product gallery. You can also use permitted material in product content, CMS pages, email or ads within the agreed rights. Safe video embedding depends on your editor, HTML filtering, theme and media integration. Use an appropriate media feature or a link if an embed is unsupported.

For an automatically displayed social feed, use the social feeds guide; platform permissions and creator rights still apply. Permission to reuse a creator's post does not establish that an account-feed connection can fetch it: confirm the supported source and access, or use an agreed manual asset. Check mobile performance and consent for third-party embeds. You can also test a simple share or copy-link control near relevant product information. Social sign-in is a separate account feature covered in social login.

Give creators room to explain the product in their own voice while keeping claims accurate. Compare the content's direct sales, production cost and useful licensed assets. Reuse can have value even when direct redemptions are modest, but do not invent a revenue value for that benefit.

Create creator codes and report redemptions

Use a separate cart-rule code for each collaboration, such as ANNA15. Redemption tells you which code an order used. A code can leak to coupon sites or other channels, and someone influenced by a creator may buy without using it. Use redemption as a campaign signal rather than proof of who caused the purchase.

Create the native cart rule

1. Open Catalog > Discounts and create a cart rule.
2. On Information, set a clear internal name and a non-empty code. Set Highlight to No if you do not want the shop to suggest it to visitors.
3. On Conditions, leave the customer restriction empty for a shareable audience code. Select a customer only when the code should be restricted to that account.
4. Set the validity dates, total quantity and quantity per user. Apply any supported product, cart-value or other restrictions needed for your offer.
5. On Actions, define the discount, free shipping or gift that you actually agreed. Check the effect on margin.
6. Test an eligible and an ineligible cart, and check interactions with other discounts.

Partial use controls treatment of a voucher value exceeding the cart total; it is separate from Highlight. There is no native Actions-tab switch named “Send to a customer”. Keep a record of the rule ID and campaign dates so later code edits do not erase your reporting context.

Export order-level redemptions

The standard Orders grid does not provide an applied-voucher filter. The Discounts list is not a creator revenue report. Use a suitable reporting extension, an export or a read-only SQL report. Under Advanced Parameters > Database > SQL Manager, the following example selects the orders using the current code ANNA15:

SELECT
  o.id_order,
  o.reference,
  o.date_add,
  o.id_shop,
  o.id_currency,
  o.current_state,
  o.valid,
  ocr.id_cart_rule,
  cr.code,
  ocr.name AS saved_rule_name,
  ocr.value_tax_excl AS discount_tax_excl,
  o.total_paid_tax_excl AS order_total_tax_excl
FROM ps_order_cart_rule AS ocr
JOIN ps_orders AS o ON o.id_order = ocr.id_order
LEFT JOIN ps_cart_rule AS cr ON cr.id_cart_rule = ocr.id_cart_rule
WHERE cr.code = 'ANNA15'
  AND ocr.deleted = 0
ORDER BY o.date_add, o.id_order

Use your installation's table prefix in place of ps_. Add your campaign date and shop filters before aggregating. If the code was changed or the rule removed, filter by the recorded ocr.id_cart_rule instead of the current cr.code. The saved ocr.name is a rule name, not necessarily its code.

Review order states, payments and refunds. The order's valid flag reflects its loggable status and is not a settlement or refund ledger. Keep separate counts for redeemed orders, discount amounts and order totals. Group money by currency and shop, and count distinct order IDs when combining rows. An order using several rules must not contribute its full revenue several times to a combined campaign total.

Pair redemptions with analytics rather than replacing one with the other. Keep utm_medium=influencer consistent if that is your chosen convention, and use utm_content to identify Anna or another creator. Count code-attributed orders as a campaign signal, not as a guaranteed minimum of incremental sales.

Reconcile results and decide what to continue

Compare reports over the same dates, currency, timezone and order policy. Record the attribution model, lookback window and whether a platform includes view-through or engaged-view credit. Advertising platforms, GA4 and your order records need not report the same number.

Calculate an economic limit

Calculate contribution after product cost, fulfilment, shipping subsidy, payment fees, discounts and expected returns, before advertising. Keep the tax treatment consistent with the revenue used in the ratio. That contribution gives you a starting limit for acquisition spend; repeat-order value needs a separate, measured assumption.

For a simplified example, a 40% contribution margin before advertising gives a break-even revenue ROAS of 1 / 0.40 = 2.5. This is arithmetic for those assumptions, not a target for every store. Include creator fees, gifted-product costs, postage, content production and staff time when reviewing the whole campaign.

Work through Anna's creator campaign

Anna receives a gifted product, publishes an agreed demonstration and gives her audience the tested ANNA15 code and tagged link. First export the qualifying redeemed orders, remove cancelled orders and apply your refund policy. Calculate contribution from the retained sales after the audience discount and the variable order costs above. If the order revenue is already net of that discount, do not subtract it again.

Campaign contribution = order revenue after discounts and refunds, excluding tax
  - product, fulfilment and delivery costs
  - payment fees
  - advertising spend, creator fees and agreed commissions
  - gifted stock at cost and its delivery
  - content production and campaign staff time

Include customer-paid delivery in the revenue above when you deduct the full delivery cost. Use consistent tax treatment and count each cost once, never in both the per-order margin and the campaign total. Keep the gift's cost separate from the sold units, rather than valuing the gift at its retail selling price. Review orders without Anna's code that your analytics or survey associates with the campaign separately, matching overlaps before forming a combined observed total. That remains an attribution view, not a calculation of incremental profit.

If the code produces orders but the campaign contribution is negative, inspect the discount, fulfilment cost and collaboration terms before renewing. If the licensed footage is useful despite limited sales, make a separate decision about buying content assets; do not turn an unmeasured future benefit into revenue in this calculation.

Reconcile overlapping signals by order

1. Start with unique order IDs and the value definition you have chosen.
2. Add available analytics transaction IDs, code redemptions and survey answers to those orders using permitted reporting tools.
3. Mark which signals overlap. Count a qualifying order once in a combined total.
4. Keep unmatched or aggregate platform figures separate when an order-level match is unavailable.
5. Separate attributed revenue, customer-reported discovery and experimentally measured additional revenue.

Browser/server deduplication solves duplicate copies of a tracking event. It does not remove overlap between analytics, voucher reports, survey answers or competing advertising platforms.

Continue a channel when its measured contribution, qualified pipeline or agreed content value justifies the cost. Adjust one part of the test when you can identify a problem. Stop when the economics or workload no longer fit. Choose those decisions from your results rather than a universal conversion rate, ROAS multiple or calendar deadline.

Locate the fault before changing the campaign

Questions owners ask

Can I start without paid ads or a tracking suite?

Yes. Choose a sustainable content test, use tagged destinations with your configured analytics and keep an order-level record of any campaign codes. Add a specific advertising integration when you have a defined use for it and can test consent and event delivery. Record production and reply time even when no media budget is spent.

Should I buy an all-in-one module when I already have working tags?

List the tasks your current modules cover and the gap you need to close. Consolidation can help if it removes repeated administration, but agree which sender will own each event before switching. Compare the proposed setup against the same rehearsal cases, including features you already rely on.

Can a feed or pixel module run the ads for me?

The connection supplies posts, product data or shopping signals according to its purpose. You or your agency still manage advertising objectives, audiences, creative, budgets and reviews in the platform. Agree who handles rejected items, expired connections and reporting faults as well as campaign work.

What result should I expect from a small first campaign?

A usable result can be evidence about a product explanation, a qualified inquiry or a costed set of retained orders. Define that result before starting and review it over your buying cycle. If the sample cannot support a profitability decision, report the uncertainty and decide whether you can afford to gather more evidence.

When should I pause rather than give the campaign longer?

Pause when the offer is wrong, the measured orders cannot be trusted, the agreed loss limit is reached or your team cannot handle the promised service. A longer run will not correct a broken destination or duplicate purchases. Resume after the cause is resolved and the relevant checks pass.

What should I receive when an agency hands the setup back?

Keep business-account access, a list of integrations and their owners, the naming and purchase-reporting rules, test evidence, connection-renewal instructions and a reversal plan. Your staff should be able to find a failed feed or event delivery and know who is responsible without depending on the original developer's personal login.