One Fexa AI credit equals one full optimisation (title, meta description, short and long descriptions, ALT tags, FAQ and structured data); 5 credits equal one translation per target language. So the smart move isn't to generate everywhere — it's to generate where it pays off. The good news: scanning and auditing are free and unlimited, which lets you analyse everything before spending a single credit.
In short: audit everything (free), prioritise high-potential pages (Google Search Console opportunities + low scores), enrich data-poor products before generating, translate only your active catalogue, and re-generate sparingly. That's how a credit delivers the most SEO return.
First rule: audit everything, generate only what counts
Scanning your store's SEO (products, categories, CMS pages) consumes no credits and stays unlimited on every plan. Use it: run a complete SEO audit of the entire catalogue before any generation.
The audit gives you, for free, the map of what truly deserves a credit:
- the SEO score of each page (low scores = big potential gains);
- the Google Search Console opportunities (pages that already get impressions but convert poorly);
- the data-poor products the AI can't write well as they stand.
Scanning is free: a credit should never go to a "random" page. Let the audit decide for you.
Lever 1 — Target Google Search Console opportunities
The most profitable pages to optimise are those that already have visibility but underperform. Your dashboard surfaces them automatically once Search Console is connected.
Low CTR (priority no. 1)
A page that gets lots of impressions but few clicks almost always signals an unappealing Title and meta description. Fexa AI flags these products (above 100 impressions with a CTR under 2%) and puts them at the top of your priorities. Fixing them costs 1 credit and acts directly on the click-through rate, without touching the ranking. It's the best gain-per-credit ratio.
Full method in Fix a low CTR.
Striking distance (close to page 1)
Keywords ranking on page 2 (positions 11 to 20) are within reach of the first page: an editorial nudge is often enough to tip them over. Reserving your credits for these pages, rather than invisible content, maximises impact. See Identify striking-distance keywords.
Lever 2 — Tackle low scores first
The dashboard's Priorities table lists the pages with the lowest scores: these are where a 1-credit optimisation gains the most points. A page that already scores well, by contrast, has almost nothing to gain — no need to spend a credit on it.
| Scenario | Spend a credit? |
|---|---|
| Very low-scoring page (thin content, missing metas) | Yes — high expected gain |
| High-impressions, low-CTR page | Yes — immediate click gain |
| Keyword in position 11–20 | Yes — tips over to page 1 |
| Already well-optimised page (high score) | No — marginal gain |
| Outside the active catalogue / inactive seasonal item | No — wait until it matters |
Lever 3 — Enrich before generating (and don't waste a credit)
A credit produces content that matches the real data you provide. Fexa AI's strength is to ground each page in your catalogue: technical specs, variants, brand, Search Console keywords. If a page has no specs, no variants and no description, the AI has very little concrete material to produce differentiating content.
The dashboard surfaces exactly these pages via the "To enrich" card (products with thin content, or no specs and no variants). The right reflex:
- First complete the product in PrestaShop (features, combinations, a few lines of description);
- Then spend your credit — the AI now has real material to ground the copy in.
Enriching a data-poor product before generating it avoids "burning" a credit on content you'll want to redo later. One well-prepared credit is worth two badly placed ones.
Lever 4 — Optimise in bulk, but by priority
Once your priorities are identified, batch generation applies the same care to hundreds of pages in one click. The key is to select by priority rather than ticking everything at once: start with low CTR and low scores, measure, then continue.
See Generate optimisations in batch for selection and launch.
Lever 5 — Translate only the active catalogue
Translation costs 5 credits per page and per language: it's the line item that drains credits fastest. A few simple rules:
- "Translate first, optimise later" is the wrong approach: optimise the page in the source language, then translate the finished version — you pay for translation only once.
- Translate only active products and the markets you actually serve. 500 products × 2 languages = 5,000 credits: make sure it's the catalogue that sells.
- Preserve your already-indexed URLs on a re-translation (option on by default) so you don't regenerate needlessly.
Full details in Translate your catalogue.
Lever 6 — Re-generate sparingly
Each new generation of an already-optimised page costs 1 credit again. Before re-running:
- use the preview, rollback and history (all free) to compare and roll back without spending;
- re-generate only if the page has genuinely changed (new data, repositioning) or if the score stays low after enrichment.
A 5-step method for an optimal credit budget
- Audit everything (free) to get scores and opportunities.
- Connect Google Search Console to surface low CTR and striking distance.
- Enrich in PrestaShop the products on the "To enrich" card.
- Generate in batch, by priority: low CTR → low scores → rest of the catalogue.
- Translate only the finished pages of your active catalogue (5 cr/language).
A monthly quota follows this logic: most stores optimise their catalogue in a few campaigns, then only spend on new arrivals and high-potential pages.
Frequently asked questions
Does scanning use credits? No. Scanning and SEO auditing are free and unlimited on every plan — that's precisely what lets you target your credits.
Where should I start to get the most from my credits? With Search Console opportunities (low CTR, keywords in position 11–20) and the lowest-scoring pages: these are the 1-credit optimisations with the highest return.
Why enrich a page before generating it? Because the AI grounds content in your real data. A page with no specs, variants or description gives little material: enrich it first so your credit produces genuinely differentiating content.
How do I avoid blowing my budget on translation? Translate only your active catalogue, and only pages already optimised in the source language. At 5 credits per page and per language, it's the most expensive line item.
Should I re-generate regularly? No. Re-generate only after a real change in data or an enrichment. Preview, rollback and history are free and save you needless generations.
See also
- Understand the credit system — exact pricing, per-plan quotas and Booster packs
- Identify striking-distance keywords — spot pages one notch from page 1
- Translate your catalogue — multilingual translation at the right cost
- Generate optimisations in batch — optimise by priority, in one click