Upstart Trustpilot Index: Reading the Q4 Tea Leaves
- Henry Dierkes
- 15 minutes ago
- 5 min read
If you’ve been hanging around Rebellionaire for a while, you know Henry’s Upstart Trustpilot Index (UTPI) has turned into one of our favorite “weird but useful” data sets. It’s not GAAP. It’s not official guidance. It’s people leaving reviews on the internet—and yet, quarter after quarter, it has mapped surprisingly well to what Upstart actually reports.
Henry just dropped the latest UTPI + UpstartIQ update (you’ll find the full PDF embedded at the bottom of this post), and there’s a lot to unpack about where Upstart might be heading into Q4 2025 earnings and early 2026.
What is UTPI again, and why should investors care?
UTPI is Henry’s running tracker of Upstart reviews on Trustpilot. The core idea:
Every funded loan can generate a review.
Reviews don’t map 1:1 to originations, but over time the relationship is stable enough to be useful.
If you track the run-rate of reviews and adjust for the “review rate” (reviews as a % of loans), you can back into a rough origination and revenue picture.
In the PDF, the UTPI chart shows the 28-day moving average grinding higher across 2025, with a big bump into year-end before holiday seasonality kicks in. At December 31, the 28-day average sits at 31.86 reviews per day—up 70% from the start of 2025 and about 10% higher than at the start of Q4. The series actually hit an all-time high of 36.04 on December 12, which works out to ~93% growth since the beginning of the year.
That’s not the story of a platform stuck in neutral.
Q4 2025: What the review counts are saying
Let’s start with the basic scoreboard:
Q3 2024: 1,249 reviews
Q4 2024: 1,696 (+36% QoQ)
Q1 2025: 1,765 (+4%)
Q2 2025: 2,423 (+37%)
Q3 2025: 2,600 (+7%)
Q4 2025: 3,054 (+17.5%)
Henry’s base-case estimate for Q4 was 3,075 reviews. Actual came in at 3,054—within half a percent of that base case. That’s not just a flex for the model; it means the same UTPI framework that’s been tracking reality pretty closely is now feeding the new UpstartIQ revenue scenario in the back half of the report.
Also worth noting: some of the slowdown into the last two weeks of December looks like pure calendar noise. Holidays drag review activity lower almost every year. What matters more is what happens as soon as the calendar flips—and the first five days of 2026 came out swinging, including new all-time highs for Friday and Sunday review totals in the UTPI era.
Funding fears, Tech CU, and Dave’s “wrong”
The other half of the Upstart story isn’t just volume. It’s funding and macro.
In mid-December, Henry ran a poll in the Upstart investor community on X asking whether Upstart would manage to secure funding for new products like Auto and HELOC in Q4. A majority—just under 53%—voted no or “partial funding.” For a bullish-leaning crowd, that’s a pretty pessimistic read on funding progress.
Two things then happened in quick succession:
Tech CU partnership. On December 17, Upstart announced a new partner—Tech CU, a credit union with more than $5B in assets. It’s the second-largest partner signed since 2024, and Tech CU is expected to fund Auto Refinance loans on the platform in 2026. Translation: a real depository funding source for a newer product line, not just personal loans.
Dave Girouard pushes back. When Henry suggested that funding was the main bottleneck for the stock, CEO Dave Girouard jumped in and replied “wrong” within minutes, then followed up with a longer explanation the next day. His framing: the primary bottleneck is economic acquisition of borrowers—which depends on conversion rate, macro risk (UMI), funding costs, automation, and acquisition strategy all working together. No single variable, including funding, gets to wear the full “bottleneck” crown.
You can read that two ways:
Funding isn’t solved, but it’s no longer the fire everyone thinks it is.
The real game is shifting toward how aggressively Upstart can open the credit box as macro improves and models get better.
UMI is quietly moving the right way
The Upstart Macro Index (UMI) chart in the PDF shows something the market hasn’t fully internalized: the macro input to Upstart’s models is easing off the highs.
November’s UMI print is 1.50.
Prior months were revised down—October to 1.43, September to 1.49, August to 1.53.
When UMI spikes, Upstart tightens its credit box, originations fall, and guidance takes a hit. That’s what happened in Q3 when a macro shock forced the company to pull back. When UMI falls, the opposite dynamic kicks in:
The credit box expands.
Loans booked at earlier, higher UMI levels start to over-perform because they were priced with extra macro “padding.”
Henry’s base case is that UMI continues drifting lower into early 2026 as rate cuts flow through and consumption cools a bit without breaking employment. That’s the macro backdrop sitting underneath the Q4 UTPI strength.
The UpstartIQ baseline: a path to beating guidance
Using the final Q4 review count, Henry plugs UTPI into his UpstartIQ model and gets to a baseline revenue scenario that looks roughly like this:
Non-core revenue: about $70.8M
Net interest income: $26M (assumes they simply meet guidance)
Servicing platform revenue: $44.8M, up around 7.5% QoQ
Core revenue: about $230.1M
3,054 Trustpilot reviews (+17.5% QoQ)
Review rate inching up to 0.63%
Revenue per loan slipping ~6.5% to roughly $474 as SDL and other newer products grow their mix
Total revenue: about $300.9M vs $288M guidance
He’s not assuming heroics here—just modest servicing growth, a slightly higher review rate (which actually hurts the revenue estimate), and lower revenue per loan as the product mix shifts toward smaller-ticket, more competitive verticals.
In his prior UTPI update, Henry said there was “greater than a 50% chance” Upstart would clear $300M in Q4 revenue. With actual UTPI data in hand, his baseline still sits on the north side of that line. If management’s $288M guide turns out to be conservative, we may look back and call this quarter a classic case of sandbagging.
What this doesn’t mean
A quick reality check before anyone gets carried away:
UTPI and UpstartIQ are frameworks, not oracles. Small changes in review behavior, mix, or review rate can move the estimates around.
None of this is a guarantee about what Upstart will print on earnings day—or how the stock will react.
Rebellionaire, Halter Ferguson Financial, Henry, and some of our clients hold UPST and can trade it at any time. That’s a built-in bias you should always keep in mind.
Use this work as one input among many, not a substitute for your own research.
Download the full UTPI / UpstartIQ report
This post only scratches the surface of Henry’s January 6, 2026 UTPI update. The full PDF walks through the methodology behind UTPI and UpstartIQ, the history of the model’s accuracy, and additional revenue scenarios beyond the baseline outlined here.
You can download the complete “Upstart Trustpilot Index (UTPI) 1/6/2026” report directly from this page and dig into the charts, tables, and assumptions for yourself.
Resources
Dierkes, Henry. Upstart Trustpilot Index (UTPI) 1/6/2026. Rebellionaire / Halter Ferguson Financial, 6 Jan. 2026.
Comments