The tracking model is fundamentally different
E-commerce optimises to a Purchase event with revenue attached. The signal is dense, fast, and quantifiable. A campaign either drives purchases or it doesn’t, and the ROAS is calculable within days.
Lead gen optimises to a Lead event with no immediate revenue. The signal is sparse, slow, and qualitative. A campaign drives leads, and whether those leads close into revenue takes weeks or months to determine. Your ad platform never sees that closure unless you wire it back deliberately.
This is why lead gen accounts that “look profitable” on Meta or Google often aren’t — the platform optimises to the wrong end of the funnel.
The campaign structure splits
E-commerce structure: Advantage+ Shopping or PMax does the heavy lifting, supplemented by retargeting and brand defence. Top of funnel is broad and creative-led, bottom of funnel is catalogue-driven.
Lead gen structure: Manual Search campaigns on high-intent keywords, lead form ads on Meta with stricter audience targeting, and aggressive remarketing. Performance Max generally underperforms for lead gen — it doesn’t have enough signal to optimise on.
The content difference nobody talks about
E-commerce content sells the product or the lifestyle. Picture-led, fast cuts, clear product visibility, social proof from buyers. The user makes the buying decision in the same session.
Lead gen content sells the conversation. Educational, problem-aware, framework-led. The user is signing up to learn more, not to buy. Asking them for a sale-style commitment too early kills the form fill rate.
Putting e-commerce-style ads on a lead gen account produces low CPMs and low CPLs and zero qualified leads — because everyone who clicks expected to buy something, not have a sales call.
The CRM is mandatory for lead gen
For e-commerce, your store backend is your source of truth — Shopify, WooCommerce, whatever. Revenue lives there, and tying it back to ads is mostly a tracking exercise.
For lead gen, the CRM is the source of truth. Without it, you can’t measure what matters: lead-to-meeting rate, meeting-to-proposal rate, proposal-to-close rate, time-to-close, and average deal size by source. These are the actual ROAS inputs. Without them, you’re optimising to a Lead count that may or may not correlate with revenue.
The platforms that need to be wired together: ads, CRM (HubSpot, Pipedrive, Salesforce), and offline conversion uploads back to Meta and Google. When a lead becomes a customer in your CRM, that data should flow back to the ad platforms within 24 hours.
The budget logic differs
E-commerce can scale aggressively and quickly — if today’s ROAS is 4×, doubling spend tomorrow usually keeps it above 3×. The signal density supports rapid scaling.
Lead gen scales more slowly. Doubling spend doubles leads, but lead quality often degrades and your sales team gets overwhelmed. The bottleneck moves from ad spend to sales capacity. The growth model is “scale spend until lead quality breaks, then fix the funnel, then scale again.”
The metrics that matter, by model
E-commerce: ROAS, AOV, CPA, repeat purchase rate, contribution margin after ads.
Lead gen: CPL, lead quality score, MQL-to-SQL rate, SQL-to-customer rate, customer LTV, payback period.
Don’t try to run a lead gen account on ROAS dashboards or an e-commerce account on lead-quality dashboards. They measure different realities.