Most ad optimization stops at the landing page. But the visitor’s journey doesn’t — it continues through signup, verification, onboarding, and first value, and every one of those steps leaks. If you’re only measuring click-to-signup, you’re optimizing the first third of a flow and paying for the other two.
Here’s how to map the whole path, find the leak, and build a flow with as few steps as it can survive.
Map the actual steps first
Write down every discrete action between “sees ad” and “gets value.” A typical unoptimized flow:
- Sees ad → 2. Clicks → 3. Lands on homepage → 4. Navigates to pricing → 5. Picks a plan → 6. Signup form (6 fields) → 7. Email verification → 8. Onboarding wizard (4 screens) → 9. Empty dashboard → 10. Figures out what to do → 11. First value
Eleven steps, each with a drop-off rate. If each step retains 70% — generous — you’re delivering about 2% of your clicks to first value. That’s the real conversion rate you’re paying for, and it explains most “our ads don’t work” conclusions.
The flow that works
The same journey, compressed:
- Sees ad → 2. Clicks → 3. Dedicated landing page matching the ad’s promise → 4. Signup (email + password, or one OAuth click) → 5. Straight into a pre-populated product state → 6. First value
Six steps. Same product, same ad. The difference is entirely in what you removed.
What to remove, in order of impact
Remove the homepage from the path. Sending ad traffic to a homepage forces the visitor to re-find what the ad promised. One landing page per ad group, headline echoing the ad (how to optimize a landing page for conversions).
Remove plan selection before signup. Asking someone to choose a tier before they’ve used anything is asking for a decision they can’t make. Let them in, then upgrade later.
Remove the credit card, if your model allows. On cold ad traffic this is often the single biggest step change in the whole flow (free plan vs free trial).
Remove form fields. Email and password. Company size, role, and phone number can be asked later, in-product, when the user has a reason to answer.
Remove blocking email verification. Let people in first and verify in the background, unless your threat model genuinely requires otherwise. A verification wall between signup and product is one of the most expensive steps in a typical flow, because the user has to leave your product to continue.
Remove the empty state. The most under-appreciated leak. A user who reaches an empty dashboard has completed signup and still gotten nothing. Seed sample data, pre-fill from what you already know, or make the first action a single obvious button.
Remove onboarding screens that teach instead of doing. A four-screen tour before the product is four chances to close the tab. Teach inside the first real action.
Where to actually look for your leak
Don’t guess which step is worst — instrument it:
- Track each step as an event, not just signup. You need a funnel from click to activation, and most accounts don’t have one.
- Segment by source. The flow that works for search traffic often fails for cold social, because intent differs.
- Watch the mobile numbers separately. Signup forms break on phones constantly, and the desktop average hides it.
- Find the biggest single drop, fix only that, and measure again. Fixing three things at once teaches you nothing about which one worked.
Feed the right event back to the ad platform
A flow fix is wasted if the ad platform is still optimizing for the wrong thing. If you report “signup” as the conversion, the algorithm will find you people who sign up and never activate — and it will get very good at it.
Send the down-funnel event: activation, or paid conversion. Fewer conversions reported, dramatically better ones. Pair it with values so the platform optimizes for revenue rather than volume (more on ad measurement).
The friction worth keeping
Not all friction is bad. Some of it qualifies:
- A card requirement filters for intent when your support cost per user is high.
- A short qualifying question can route users to the right onboarding path, if it genuinely changes what they see next.
- A demo booking is correct when contract values are large enough to justify a human conversation.
The test: does this step make the next step better, or does it just make you feel better? Keep the first kind, delete the second.
Why referred traffic makes all of this easier
Worth noting when you compare channels: a visitor who arrives from a trusted recommendation tolerates more friction and converts better at every step of the same flow, because the persuasion happened before the click. That’s a large part of why referral and affiliate traffic tends to show a better cost per activated user than cold paid traffic on identical pages — and why the two are usually better run together than as alternatives (paid vs organic acquisition).
FAQ
How many steps should a SaaS signup flow have?
As few as the product can survive — realistically ad click, landing page, signup, and then a product state that already has something in it. Each additional step compounds drop-off across the whole funnel.
Should I require email verification before letting users in?
Usually not immediately. Verifying in the background while the user starts using the product avoids sending them out of your app at the moment of highest motivation, unless abuse risk requires otherwise.
What conversion event should I send back to ad platforms?
The furthest-down-funnel event you can reliably measure — activation or paid conversion rather than signup — so the platform optimizes for customers instead of form fills.
More conversion strategy is in the growth & pricing hub.