
Unlock the power of transaction data.
Explore the risks and possibilities with a prompt for ChatGPT, Claude, or your agent.
Second Measure turned card transactions into daily estimates of company performance. Founded in 2015 by Michael Babineau and Lillian Chou, the Summer 2015 YC company normalized billions of purchases into revenue growth, customer growth, retention, cohorts, market penetration, and competitive benchmarks.[1][2]
The company did not fail. It expanded from investors to consumer brands, reached tens of millions in founder-reported annual recurring revenue, and was acquired by Bloomberg on December 24, 2020.[1][3] The acquisition mechanism was data distribution and context: Second Measure supplied normalized alternative data; Bloomberg supplied company identifiers, fundamentals, institutional reach, entitlements, and established research workflows.
Babineau and Chou founded Second Measure around proprietary consumer-purchase feeds. Their thesis was that transaction data could reveal public and private company performance days or months before conventional disclosures.[2]
At launch, Babineau said the company was focused “primarily towards VCs and hedge funds.” Hedge funds used the data to “inform their financial models.”[2] Venture firms could inspect private-company momentum; public-market investors could compare reported narratives with observed consumer spend.
The difficult work sat between raw transactions and an investable conclusion. Merchant descriptors were inconsistent. A consumer panel could differ from the population. Small customer bases could create false affinity signals. Second Measure had to normalize merchants, model panel bias, distinguish cohorts, and expose uncertainty without making the product unusable.
That discipline let the company widen beyond finance. By 2018, customers included Spotify, Barneys, and Blue Apron alongside investment firms.[4] Brands wanted the same evidence for market share, retention, and competitor benchmarking.
Second Measure analyzed billions of anonymized purchases obtained through proprietary data-provider relationships. Its interface showed estimated revenue and customer growth, retention, cohorts, market penetration, and competition for thousands of companies.[2]
Merchant normalization was central. A single company could appear under inconsistent descriptors across processors, locations, or products. Resolving those rows into a coherent company and brand hierarchy determined whether the dashboard represented economic activity or noise.
Panel correction mattered as much. The Company Affinity feature used Bayesian methods because naive estimates exaggerated relationships involving merchants with small customer bases.[4] This illustrates the broader product: raw counts became useful only after statistical adjustment and defensible comparison.
Bloomberg later connected the dataset to traditional market information. Its ALTD function supported discovery and analysis of alternative data; Data License exposed feeds for enterprise and quantitative workflows.[6][7] The current product advertises more than eight years of history with two-, three-, or seven-day delivery lags.[8]
Second Measure began with venture firms and hedge funds, then expanded to corporate strategy and consumer-insights teams. Reported users included Goldman Sachs, Spotify, Instacart, Postmates, and Domino's.[9]
No audited market size, contract value, renewal, or margin data was observed. Babineau's current YC profile says the company reached tens of millions in annual recurring revenue and a nine-figure Bloomberg exit; both are founder self-reported.[1]
Bloomberg's 2024 release described more than 20 million U.S. consumers, 3,000 public and private companies, over 4,000 brands, billions of transactions, and a three-day lag.[7] These are Bloomberg-reported figures.
Second Measure competed with other alternative-data vendors, internal data-science teams, market-research panels, and conventional company research. Its advantage was not mere access to card rows. It was merchant resolution, panel adjustment, daily delivery, and a usable analytical layer.
Bloomberg strengthened each dimension through institutional distribution and reference data. It could connect an alternative signal to identifiers, filings, estimates, news, and portfolio workflows. That made Second Measure's dataset more valuable without changing its underlying thesis.
Second Measure sold subscription access and data products to investors and consumer brands. The value came from proprietary feeds, normalization, statistical methods, and low-lag company analytics.
The company reportedly raised about $25 million before acquisition, though the exact chronology was not confirmed from primary financing documents.[5] YC names Bessemer, Goldman Sachs, Citi Ventures, and YC among investors.[1]
No audited pricing, gross margin, contract term, renewal, acquisition cost, or data-licensing cost was found. The founder-reported revenue and exit figures suggest material scale, but cannot substitute for audited economics.
By 2018, Second Measure employed 51 people and served investment firms plus brands such as Spotify, Barneys, and Blue Apron.[4] Later reporting named Instacart, Postmates, Domino's, Spotify, and Goldman Sachs as users.[9]
Bloomberg offered a concrete investor example: Second Measure data indicated Netflix subscriber gains around password-sharing changes before Netflix reported results.[6] The 2024 Data License expansion shows continued product investment and distribution beyond the Terminal.
Bloomberg acquired Second Measure to add daily transaction analytics to its fundamentals and market-data products.[3] Babineau and Chou pointed to Bloomberg's “leadership, scale, and complementary products” as the reason the combination could accelerate client insights.[3]
The mechanism was context plus distribution. Second Measure could sell a normalized signal. Bloomberg could put that signal beside company fundamentals, estimates, news, and portfolio tools already used by institutions. Its entitlement and delivery infrastructure also opened enterprise and quantitative channels.
Post-acquisition work validates the thesis. Bloomberg combined the teams to build ALTD and later distributed Second Measure through Data License.[6][7] The product survived as a proprietary Bloomberg dataset.
A founder-reported nine-figure exit and tens of millions in recurring revenue imply strength, not rescue.[1] Second Measure might have remained a major standalone alternative-data vendor with its own enterprise distribution.
The evidence cannot rank acquisition against independence. Terms, margins, renewal, investor returns, and retention packages remain unknown. No demand collapse or operational crisis was observed. The careful conclusion is that Bloomberg could monetize and contextualize the signal more broadly, not that Second Measure lacked an independent future.
The product relied on anonymized and aggregated purchase data, licensed provider relationships, merchant normalization, and panel representativeness. No regulator action or breach appeared in the observed evidence. Neither did a full sampling audit, re-identification analysis, or provider list.
That opacity is material. Alternative-data quality and legitimacy depend on who participates, what they consented to, which uses are permitted, and whether aggregated results protect individuals. A modern rebuild should make that bargain explicit rather than treating the panel as invisible infrastructure.