Doubling Signup and Login Completion Rate
Awesome Free is a new consumer rewards platform. It doesn't run its own accounts. It shares them with Advocado CRM's loyalty program, which already has more than 2 million users.
That shared setup shaped every decision. Whatever I changed about authentication still had to work for 2 million members already living on the existing account system.
So a failed login wasn't a casual bounce. It was a ready-to-redeem customer walking away at the last step.
Password friction
Returning users kept forgetting their passwords. Each reset meant another SMS charge and one more reason to give up.
Signup friction
New users hit a wall of required profile fields before they could claim a thing. Right at peak intent, the reward sat stuck behind a form.
I looked at a few directions and weighed each one on experience, engineering cost, and how well it fit the existing accounts.
My first proposal: phone + OTP and Google login.
Google accounts line up one-to-one with phone numbers, so it would clean up identity and take the password problem off the table at the same time.
Enter phone → verify OTP → done
Signup now creates a partial-registration account with nothing else required. Cutting the form to two steps attacks the friction that was stalling new users at their highest-intent moment.
Complete profile later → earn rewards
Deferring the profile didn't cost the business its data. Tying completion to a reward turned a mandatory form into an opt-in, so business keeps its data and users keep their momentum.
Partial-registration users → OTP login
They never set a password. OTP is the only method that fits, and it removes the biggest drop-off point for the fastest-growing segment.
Password users → password or OTP login
Forcing OTP on everyone would slow down people who know their pass & drive up SMS cost. Offering both keeps returning users fast and keeps verification spend in check.
Google users → Google or OTP login
Google stays the one-tap path, with OTP as a shared fallback. Every account type resolves to a method that works, so no one hits a wall.
Partway through, a request reached me secondhand: make OTP the only way to sign in. It was a big reversal, and it hadn't come straight from the person who'd actually decide.
When a big directional change shows up secondhand, build just enough to get everyone aligned, then confirm before you commit.
Authentication completion
2.4×
+138%
no backend rebuild