In partnership with
Oregon deliberately made sign-up harder and lost a quarter of the people who tried. Four months later, the harder version had produced the better result. This week: what that experiment says about the friction you may be removing from your product, and why an eight-person software company doing roughly $8 million a year replaced its trial with a free plan.
| In this issue | ||||||
|
Startups on Carta recorded 421 M&A exits in the first half of 2026, the most on record for that stretch and 16% more than a year earlier, per Carta's Data Desk. Skip what companies like yours are worth and find what buying you solves for this buyer, because two years off a roadmap, a customer base kept or a competitor blocked can be worth far more than the multiple on your last financing.
How Jennifer Aniston’s LolaVie brand grew sales 40% with CTV ads
For its first CTV campaign, Jennifer Aniston’s DTC haircare brand LolaVie had a few non-negotiables. The campaign had to be simple. It had to demonstrate measurable impact. And it had to be full-funnel.
LolaVie used Roku Ads Manager to test and optimize creatives — reaching millions of potential customers at all stages of their purchase journeys. Roku Ads Manager helped the brand convey LolaVie’s playful voice while helping drive omnichannel sales across both ecommerce and retail touchpoints.
The campaign included an Action Ad overlay that let viewers shop directly from their TVs by clicking OK on their Roku remote. This guided them to the website to buy LolaVie products.
Discover how Roku Ads Manager helped LolaVie drive big sales and customer growth with self-serve TV ads.
The DTC beauty category is crowded. To break through, Jennifer Aniston’s brand LolaVie, worked with Roku Ads Manager to easily set up, test, and optimize CTV ad creatives. The campaign helped drive a big lift in sales and customer growth, helping LolaVie break through in the crowded beauty category.
The losing version
Oregon ran a sign-up experiment that looked terrible on day one. Four months later, the losing version had produced 18% more of the behavior Oregon actually wanted.
In 2019, Oregon's Department of Transportation intentionally annoyed 13,564 commuters.
The agency's carpool website was migrating accounts, and decided to take the opportunity to conduct an A/B test. If you were among the 13,663 users in the lucky cohort, you got a simple password reset and went forward with your day.
If you were one of the unlucky 13,564, though, you had to rebuild your account across three pages: Name, email address, a new password, home and work ZIP codes, an organization if you had one (which, as a commuter, you probably did). Exactly the type of thing that makes the typical user mumble their favorite curse word before moving forward with their lives.
Not surprisingly, sign-ups in the latter group fell by a quarter.
But the group carpooled more anyway. Over the next four months the harder-sign-up cohort, counting everyone emailed, logged 5,106 trips against 4,311 for the lucky one.
Or, in SaaS-friendly jargon: Oregon lost a quarter of its conversions and gained 18% more of the behavior it wanted.
Any dashboard worth its salt would have called the three-page form a disaster on day one and been right, at least for a while. Until reality sets in, and you learn you can improve the conversion and worsen the outcome simultaneously.
The bet on free
Circleback is an eight-person meeting note-taker doing, by TechCrunch's estimates, about $8 million in annual run-rate revenue. It says it has been profitable since its 2024 funding raise. On Aug. 31, TechCrunch reported that the company had added a free plan, replacing a limited trial with an indefinite way into the product.
The logic is familiar. Co-founder Ali Haghani told TechCrunch the previously limited trial had produced a big drop-off in users. Remove the expiration, more people keep using the product, more people see it, more of them eventually pay.
"If we just open the gates and allow more people to use the product," Haghani said, "that's gonna bring Circleback in front of more people."
The company says it buys no Google or Meta ads and that the free plan will serve as its marketing expense, with the gate moved to 30 days of history, two teams and a limited API.
Maybe. Oregon is the asterisk. Removing a barrier will put more people through the gate, but getting more people through the gate and producing committed users are two different problems.
Why the harder version worked
The Oregon sign-up was an experiment, run inside the state's live account migration by Holly Dykstra, Shibeal O'Flaherty and Harvard Business School's Ashley Whillans. Half the users got a password reset; the other half rebuilt their account across three pages.
Sign-ups in the three-page group fell 1.3 percentage points from a 5.1% base, the 25% drop. Then that group took, in the authors' words, "1.6 times more carpool trips per week over four months," counting everyone emailed, and the raw totals came out 5,106 to 4,311.
The authors call it the buy-in effect.
"The target behavior feels more valuable when we've invested time in it," said Whillans, to HBS Working Knowledge.
Their forthcoming paper in Management Science names the conditions under which the effect is most likely to appear: the behavior has to require follow-through, joining has to be voluntary, the effort has to be modest, and it has to be goal-relevant, tied to the thing itself.
The effect has limits. When signing up is most of the work, as with a benefits program that renews automatically, the authors say reducing friction may still be the better strategy. And there is a point where added effort simply costs you users.
"Excessive friction," they write, "may reduce uptake without commensurately increasing commitment."
A harder trial, a different customer
There is some evidence of the same pattern in software. ChartMogul and ProductLed looked at 200 software products in January 2026 and found that free trials requiring a credit card converted 30% of users to paid, more than five times the rate of trials that did not require one. Only 20% of the free-trial products in the study asked for a card upfront.
That does not mean the credit card caused the difference. Asking for one may simply screen out people who were never likely to pay, and the report warns that the requirement can cost you sign-ups if the product has not earned the ask.
A 30% conversion rate makes the card look like the answer. But if the card only filters for people who were already going to pay, you have learned how to screen customers, not how to make them more likely to stay.
Oregon points to the harder question: what can you ask a customer to do that makes them more likely to keep going?
The value of friction
Oregon was studying carpooling, not software. But the finding raises a useful question for any trial or free plan: what should someone have to do before you count them as interested?
Ask for effort that moves them toward the reason they came. Connect the calendar. Import the data. Invite the teammate. Build the first dashboard. Each makes the product more useful while asking the customer to put something into it. Seven profile fields just make the form longer.
The harder question comes when you test removing that effort. An easier flow will usually win the metric closest to the change. More people will finish it. More people will activate. The result gets harder to read when you follow those same people for another month, or four.
Oregon could have stopped on day one. The three-page form produced 25% fewer sign-ups. Kill it.
Instead, they kept measuring.
Four months later, the losing version had produced 795 more carpool trips.
So before you celebrate the experiment that lifted conversion, keep the losing cohort around long enough to find out what you actually won.
Markets now put a September rate hike at 55%
The Fed held in July. On Aug. 28 at Jackson Hole, Chair Warsh said the Fed has "work to do" unless inflation moves to 2% "clearly and at sufficient speed," and futures price a hike on Sept. 16 as the likelier outcome (Fortune, TD Economics). Founder read: Model a hike against any debt you may refinance or raise in the next 12 months; the same hike costs most when it comes with slower growth, tighter covenants or less lender appetite.
U.S. employers plan 3.3% base-pay budget increases for 2027
Korn Ferry's June survey of 5,512 employers across 135 countries found U.S. employers plan to lift base-salary budgets by an average of 3.3% next year, with a median of 3% for all employees and for executives alike (HR Dive). Founder read: Price the jobs instead: compare each critical role's current pay with what replacing that person would cost in 2027, then put the adjustment money where that gap is largest.
Google now reports AI Overviews impressions and offers an opt-out
On Aug. 31 Search Console gave every site owner a report of impressions from AI Overviews and AI Mode, without clicks, and a switch that removes the site from both; Google says the switch leaves ordinary rankings alone (Search Engine Land). Founder read: Before you touch the opt-out, find the pages behind those impressions; if Google keeps choosing your site to answer what buyers ask before they buy, those buyers may be reading you weeks before any deal shows up as search-sourced.
Introducing The First Agentic CRM
Get revenue agents, workflows, and automations across every stage of your motion. Access customer data in real time through Attio's web app, MCP, API, and SDK.
Then Ask Attio anything about your business and get instant answers.
It's the CRM that runs the work behind every win.
Find one onboarding step you have considered removing because it hurts conversion. Before you touch it, see what happened to the customers who completed it.
If those customers retain better or produce more revenue months later, the step may be costing you signups while improving the customers who remain. That does not tell you to keep it. It tells you what the conversion metric leaves out.
Definition of done: You know whether completing that step is associated with better retention or more revenue per signup.
Level of effort: About two hours if the data is already available.
Thanks for reading. See you next Wednesday.
— Jason
P.S. Hit reply and tell me: what does a new sign-up have to do before you count them as a customer?
