B2B Paid Search and New-Customer Attribution for EIS
Finding the new customers hidden inside paid-search revenue
TL;DR
- Problem
- Brand search looked very profitable, but repeat buyers and first-time customers were counted together.
- What we did
- We matched paid-search visits to order history to separate first orders from repeat purchases.
- Result
- New-customer orders linked to paid search rose from 22 to 83 across matching April–July periods, and their revenue rose about 70%.
- Industry
- Industrial distribution
- Timeline
- April–July 2026, compared with 2025
- Engagement
- Client work delivered through Aux Insights
- Tools & channels
- Google AdsGA4ERP order history
3.8×
Monthly new-customer orders from paid search
+70%
Revenue from those orders
8% to 17%
Paid search share of new-customer orders
Strong brand-search returns hid repeat purchases
EIS’s brand-search campaign reported a 78× return. But a returning customer searching the company name counted the same as a first-time buyer, so leadership couldn’t see which campaigns actually brought in new customers before deciding where to spend.
Separate first orders from repeat business
We matched visits from paid search to order records and used purchase history to separate first orders from repeat business. That showed how brand searches and product-category searches each contributed to acquisition.
When brand spending was reduced, organic revenue rose by roughly what brand-paid revenue lost, which gave the budget discussion something concrete to work from.
Same four months, one year apart
| New-customer measure | 2025 | 2026 |
|---|---|---|
| Orders linked to paid visits | 22 | 83 |
| Monthly average | 5.5 | 20.8 |
| Non-brand orders | 12 | 53 |
| Paid search share of new-customer orders | ~8% | ~17% |
New customers became visible on their own
Most of the order growth came from non-brand searches, the product and category terms people use before they know the company. Brand search still drove most new-customer revenue.
Existing customers remained the bulk of the business, 83% of web orders in July 2026. What changed is that acquisition could be seen and budgeted on its own instead of being folded into repeat revenue.
How to read these numbers
Orders are linked to paid visits through purchase history. That shows association, not proof that an ad caused each order. The 2025 starting point was small (22 orders), so the percentages look larger than the volumes behind them. The brand-spend observation was not a controlled test.
Services used:Paid advertising managementMarketing analytics, attribution and reportingData pipelines and warehouses
