Google Ads in Morocco — Search, Performance Max, Shopping, YouTube under one architecture.

Below 30 weekly conversions consolidated server-side, start on Manual CPC or Maximize Clicks with a per-keyword manual CPC cap — Smart Bidding doesn't have enough signal to learn. Between 30 and 100 conv/wk, switch to tCPA with a target 15% above historical CPL to give the algorithm headroom. Above 100 conv/wk and with conversion value fed in (GA4 Enhanced Ecommerce), move to tROAS for e-commerce. The learning phase lasts 7 to 14 days after each strategy change — during that window, don't touch anything else.

Exact + phrase match Search for the launch in 80% of cases. Performance Max needs at least 50 confirmed server-side conversions over 30 days for the algorithm to break out of trial-and-error. Launched too early, it sweeps the inventory at random and burns 40 to 60% of the budget on irrelevant YouTube and Display placements. The standard Webotic sequence: months 1-2 on exact + Branded + phrase match Search, feeding the account with confirmed server-side conversions, then a progressive shift to PMax in month 3 with asset groups segmented by product category or persona. For e-commerce operators with a clean Merchant Center catalog, Shopping Smart can be activated in month 2 alongside Search.

The pragmatic floor sits at MAD 5,000/month in media spend for an exact-match Search campaign to accumulate exploitable signal (≥ 30 weekly conversions in lead-gen, ≥ 50 in e-commerce). Below that, the algorithm thrashes and performance swings ±50% week over week. For Performance Max and Shopping, the threshold rises to MAD 10,000/month because those formats need more signal to exploit inventory. Adding Webotic fees (MAD 6,500/month on the single-format pack), the real entry ticket lands at MAD 11,500/month in steady state, plus the initial MAD 8,000 setup.

Three causes in 90% of cases. First cause: degraded Merchant Center feed — products without GTIN, images under 1200×1200, Google taxonomy category badly mapped, brand attribute missing. On audited accounts, 9 out of 14 have more than 30% of products disapproved or under warning. Second cause: incomplete asset groups — without 5 headlines + 5 descriptions + 15 images + 5 videos per asset group, Google auto-generates thin content that doesn't engage. Third cause: missing audience signals — without customer match and in-market lists as input, the algorithm starts from zero on every launch. Fixing these three axes moves the median PMax ROAS from 1.8x to 4.2x on Webotic Moroccan e-commerce accounts.

Three technical levers combined. First lever: Server-Side GTM hosted on a client subdomain (data.example.ma), which bypasses cross-domain blocks and Safari ITP on the tracking side. Second lever: Enhanced Conversions for Web and for Leads — email, phone, and address parameters are SHA-256 hashed and sent in parallel with gtag.js, letting Google match the conversion even when the third-party cookie is dead. On Webotic accounts, Enhanced Conversions recovers 11 to 24% of additional conversions. Third lever: offline conversion import via gclid + CRM status, which sends Google the real business-value signal (meeting booked, quote signed, contract) rather than the form-fill proxy.

Three campaigns maximum for a budget under MAD 25,000/month. One Search Branded on your brand name and variants (CPC MAD 0.30-1.20, very high ROAS but limited volume). One Search Generic High-Intent on transactional commercial queries in exact + phrase match (CPC MAD 1.50-6.00 by vertical). One Performance Max launched in month 2 after accumulating 50 confirmed server-side conversions, with asset groups by product category or persona. Above MAD 25,000/month, we add Search Competitor on competitor names in exact match, then YouTube TrueView for Action targeted in-market and customer match. Excessive fragmentation (10-15 campaigns on a starter account) kills the signal — we consolidate on arrival at Webotic.