Somewhere in a grocery store, a person with a clipboard was copying prices. Somewhere else, an executive was preparing to make a very expensive decision with the results. The numbers would arrive late, flattened into a metro average, and polished until the living mess of the shelf looked orderly. Ben Reich thought this was absurd.
He and Dan Gallagher were software engineers at Applied Predictive Technologies, the Arlington analytics company later acquired by Mastercard. Their work put them close to large retailers and consumer packaged goods companies. They kept hearing versions of the same confession: even sophisticated businesses sent people up and down aisles to check prices by hand. The ritual belonged to another age, but the decisions riding on it were thoroughly modern and often worth millions.
A price is a small object with a large social life. It changes by neighborhood, competitor, promotion, package size and Tuesday afternoon. Compress it into a 12-week regional average and you may produce a respectable number that describes no store a shopper can actually visit. Reich and Gallagher saw a data-collection problem disguised as an industry custom. In 2014, they co-founded Datasembly to solve it.
The shelf, translated
How old is the picture?
Origin storyThe useful education of a Queens tinkerer
Reich grew up in Queens, New York. His mother was a nurse; his father taught English in public school. He has said he learned to code young and felt pulled toward technology, software and computer science. At Cornell University, he chose a combination that sounds eccentric until one considers his later company: mathematics and linguistics, with a minor in computer science.
Numbers supplied structure. Linguistics supplied context. As a Cornell research assistant, he investigated limits on the human capacity to understand natural language. Datasembly would eventually face a related commercial puzzle: how do you turn an unruly flood of public information into something a person can interpret and use?
After graduating in 2011, Reich joined APT. Gallagher, another Cornell alumnus, was there too. Reich has spoken about his co-founder with conspicuous admiration, calling him a gifted engineer and crediting Gallagher with suggesting they strike out together. It is a useful counterweight to the solitary-genius founder fable. Datasembly began with two colleagues hearing the same customer frustration and deciding it deserved better machinery.
“I learned to code at a young age.”Ben Reich, on his early pull toward technology
They were first-time founders. Neither knew how to raise capital or start a company. Early customers saw enough promise to pay, helping keep the venture alive while its creators learned the rest. By 2015, Datasembly appeared in 500 Startups' fourteenth batch with a pitch that fit on a shelf label: the price of every product, in every store, every day.
The productA local market refuses to sit still
Datasembly gathered publicly available product records from online and physical retail channels, then organized them into a view of price, promotion, assortment and availability. The point was not merely volume. It was specificity. A cereal brand could see whether a promotion appeared where it was planned. A retailer could watch a rival change a key price across town. A team could compare private-label assortment without dispatching an army of clipboards.
The timing became unusually favorable. Click-and-collect and local delivery made store-specific information more visible online. Then the pandemic made retail conditions wildly changeable. Shelves were rearranged for toilet paper and disinfectant. Assortments shifted. Availability fractured. A leisurely data cycle became a liability, and Datasembly was positioned to scale a more immediate one.
Capital followed the demand. In August 2020, the company announced a $10.3 million Series A led by Craft Ventures, with Valor Siren Ventures participating. In May 2023, it announced a $16 million Series B led by Noro-Moseley Partners. Reich's comment at the second raise was characteristically about evidence rather than theater: the business had kept growing in a more difficult economy because customers still needed its data and tools.
Public signalWhen the shelf became national news
A commercial data platform found itself describing public anxiety in 2022. Infant formula was disappearing from American shelves. Datasembly measured a nationwide out-of-stock rate of 43 percent for the week ending May 8, far outside the customary range its data had shown earlier. The number moved through news reports and official documents because it answered the urgent question with unusual clarity: how bad is the shortage right now?
The same machinery offered a live view of grocery inflation, including the rise in egg prices. A shopper experiences inflation at a register, not inside a national index. Datasembly could break movement down by state, category, urban or rural market, even parts of a town. Its granularity exposed an important fact: two true statements about prices can feel contradictory because they describe different places.
By 2024, Reich was describing a market past the first blunt wave of inflation. Consumers expected more restraint. Private labels added pressure. Retailers could no longer move uniformly and assume competitors would follow. Precision mattered in promotion, shelf space and margin negotiation. The old cudgel of a broad price increase was giving way to something closer to tweezers.
“It is getting about efficiency and precision.”Ben Reich on the changing retail-price contest, 2024
The dealJoining the shelf to the register
On June 24, 2025, SPINS announced that it had acquired Datasembly. The price was not disclosed. The logic was. Datasembly brought timely, store-level information about pricing, promotion and assortment. SPINS brought structured product, category and performance intelligence across consumer goods. Together, the companies could connect what appeared on the shelf with a broader understanding of what products were and how they performed.
Reich framed the transaction as an expansion of the original mission. Datasembly had been founded, he said, on the belief that transparent access to pricing data could change how brands and retailers operated. SPINS offered a route to more partners. A 2026 industry directory still listed Reich as Datasembly's chief executive and SPINS as its parent, suggesting continuity after the celebratory photographs had passed through the feed.
The combination also completed a useful sentence. Datasembly could describe what was happening to an item at a store: the price, the promotion, the gap where stock should be. SPINS could add detailed product and category context, plus a view of performance. Brands could connect shelf execution with results; retailers could compare assortment gaps, emerging regional patterns and competitive moves. The machinery remained complicated, but the customer question became pleasingly plain: what changed, where did it change, and did it matter? Reich had spent years improving the first two answers. Inside SPINS, the third could become easier to reach.
- Reich graduates from Cornell and joins Applied Predictive Technologies.
- He co-founds Datasembly with engineer Dan Gallagher.
- The company announces a $10.3 million Series A.
- Its store-level numbers track formula shortages and food inflation.
- A $16 million Series B funds the next stage of growth.
- SPINS acquires Datasembly and pairs shelf intelligence with a wider CPG data set.
What travelsThe founder lesson hiding in the workaround
There is a temptation to make Reich's story about big data. The more portable lesson is almost the opposite. He started by noticing one embarrassingly small thing: people were still walking aisles and writing down prices. The important clue was not futuristic. It was a workaround everyone had learned to tolerate.
He also matched the unit of data to the unit of action. Retail decisions happen at particular stores, against particular competitors, during finite promotional windows. A gorgeous national average may be useless if a manager must decide what to do in Oakland on Friday. Granularity was not a technical ornament. It was the product.
Three things worth stealing
- Listen for the repeated complaint that an industry delivers with a shrug.
- Measure reality at the same level where the customer has to act.
- Keep the historical record. Today's operational exhaust may explain tomorrow's crisis.
The historical record mattered. By 2023, Reich said Datasembly had accumulated more than a trillion data points. Today's price is useful; years of prices reveal volatility, seasonality and the strange choreography of competitors. A data business becomes more interesting when its memory grows faster than its dashboard.
Reich's path from Queens to Cornell, APT, Datasembly and SPINS has a quiet coherence. He learned to code early. He studied the structures beneath numbers and language. He found an industry where the available picture was too slow and too broad. Then he and Gallagher gave it a sharper vocabulary.
The grocery aisle still looks ordinary. Boxes face forward. Yellow labels interrupt white ones. A promotion ends, a competitor moves, an item vanishes. Beneath that calm surface, the shelf changes constantly. Reich built a company by insisting those changes were worth reading before they became old news.