True profit based optimisation means your bidding and budget decisions are driven by what matters after the click: closed revenue quality, contribution margin, churn risk, refunds, and actual profit.
What “Profit Based” Actually Means (Define It First)
Before you connect anything, decide what number Google Ads should learn from. For most businesses, revenue is not enough.
Common definitions that work well:
- Gross profit: Revenue minus cost of goods sold
- Contribution margin: Gross profit minus variable delivery costs (shipping, payment fees, commissions, support cost per order)
- Net profit proxy: Contribution margin minus expected refunds and chargebacks (or a return rate factor)
If you skip this step, you will upload revenue and still end up scaling campaigns that look good in Ads but hurt finance.
Why CRM Alignment Changes Bidding Quality
Google Ads bidding systems respond to the conversion values you give them. If you only send top of funnel conversions, the algorithm will optimise for volume, not quality. When you send offline outcomes like closed won revenue, margin, and deal quality signals, Google can start finding more users who resemble your best customers, not just your easiest leads.
This is the biggest shift:
You stop rewarding clicks that convert quickly and start rewarding clicks that convert profitably.
Book a free consultation with us to learn more about it.
The Data Foundation You Need (Without This, It Breaks)
1) A Reliable Click Identifier
To match CRM revenue back to the ad click, you must capture and store identifiers such as:
- GCLID (Google Click ID)
- GBRAID and WBRAID (often used in iOS and certain privacy scenarios)
These IDs must travel from your landing page to your form submission, then into your CRM record, and then into your offline conversion upload.
2) Clean Identity and Event Mapping in Your CRM
Your CRM should be able to answer these questions consistently:
- Which lead belongs to which opportunity or deal
- Which deal was influenced by Google Ads
- What stage it reached, and when
- What value should be assigned when it hits each stage (if you do stage based modelling)
The most common failure is messy lifecycle mapping, like duplicate leads, multiple opportunities per contact, or a “closed won” process that is not used consistently by the sales team.
3) A Conversion Taxonomy That Matches Your Sales Cycle
Do not upload everything as one single conversion. Build a structure Google can learn from: For lead gen, a practical setup is:
- Primary conversion: Closed Won (with profit value)
- Secondary conversions: Sales Qualified Lead, Opportunity Created, Quote Sent (optional, with weighted values)
For ecom, common setup is:
- Purchase (value adjusted for margin or refunds)
- Subscription activated (value adjusted for churn risk if applicable)
This prevents the algorithm from over learning on noisy micro events, while still giving it enough feedback during longer sales cycles.
Three Proven Technical Approaches
Approach A: Offline Conversion Import (Best for Lead Gen and High Ticket)
This is the most direct way to teach Google what actually closed. How it works:
- User clicks the ad and arrives with GCLID.
- You capture GCLID in a hidden field and store it in CRM.
- When the deal closes, you upload an offline conversion to Google Ads with:
- GCLID (or other identifier)
- Conversion action name (Closed Won, for example)
- Conversion time (important)
- Conversion value (profit or margin based value)
- You bid using value based strategies like Maximise conversion value or target ROAS.
Where it shines:
Longer sales cycles
High variance in deal value
Offline sales team involved
Approach B: Enhanced Conversions for Leads (Good When IDs Are Harder)
If you struggle with click IDs or have consent limitations, enhanced conversions for leads can help by using hashed first party data (like email or phone) to improve match rates, depending on your setup and permissions.
Use this when:
- You can reliably collect consented lead data
- You want better matching in privacy restricted environments
- You still plan to upload qualified outcomes later
Approach C: Data Warehouse and Server Side Pipelines (Best for Scale)
If you have multiple data sources, refunds, subscriptions, and product margins, you will eventually want a warehouse led approach:
- Website events and Ads clicks
- CRM outcomes
- Finance system data (refunds, discounts, COGS)
- A transformation layer that calculates profit
- A scheduled upload back to Google Ads via API
This approach is ideal when “profit” depends on more than one system and needs consistent rules.
The Part Most Teams Miss: Value Modelling During the Lag
If your sales cycle is 14 to 90 days, Google will not get fast feedback from closed deals alone. That can slow learning, especially in smaller accounts.
A smart fix is a two layer value model:
1. Early stage value (fast feedback)
Example: Sales Qualified Lead gets a value based on expected profit, not just a random number.
2. Final true value (slow but accurate)
Closed Won uploads overwrite the model with real profit.
A simple expected value formula can be:
Expected Profit = Average Profit per Closed Won × Stage to Close Probability
So if your average profit is 50,000 and SQL to close rate is 20%, your SQL value becomes 10,000. This keeps bidding aligned with profit even before deals close.
Making “Profit” Real: Margin, Refunds, Discounts, and Churn
If you only send gross revenue, you will scale campaigns that sell low-margin products aggressively and miss the best margin pockets.
Practical profit adjustments:
- Subtract COGS using product-level margin tables
- Subtract shipping and payment fees if they vary widely
- Apply a refund factor if returns are common
- For subscriptions, send value as first payment margin plus expected retention value, not just first month revenue
If you cannot calculate perfect profit, start with a strong proxy. Even a margin-weighted revenue model is a major step up from raw revenue.
Quality Control Checklist (So You Do Not Feed Google Bad Data)
Matching and Timing
- Ensure conversion time is accurate and not the upload date
- Use consistent time zones between CRM and Google Ads
- Avoid duplicate uploads for the same deal unless you intentionally support adjustments
CRM Hygiene
- Enforce clear definitions for stages
- Deduplicate leads and standardise opportunity creation rules
- Ensure only real closed won events trigger uploads
Audit and Monitoring
- Track match rate between CRM uploads and Google Ads
- Compare uploaded totals vs CRM totals weekly
- Watch for sudden value drops that indicate margin logic bugs or refund logic misfires
What You Get When It’s Done Right
When CRM revenue and profit signals are flowing back into Google Ads properly, you typically see:
- Less budget wasted on low quality leads
- Better scaling because bidding is trained on the right customers
- Clearer performance reporting that finance trusts
- More stable growth because profit protects you from “fake ROAS” spikes
Book a Call With a Y77.ai Expert
Want to set this up the right way, without messy tracking gaps or unreliable CRM matching?
Book a call with a Y77.ai performance expert and we will review your current CRM, conversion tracking, and Google Ads structure, then map a clean profit based optimisation plan you can actually run. You will walk away with a clear data flow blueprint, conversion taxonomy, and the exact fields and upload method needed to send true profit back to Google Ads.
Book a free consultation with us.