Failure intelligence, not failure trivia Monday, July 27, 2026
F Failurepedia

Social Commerce

Amazon Spark

Amazon Spark was Amazon's Instagram, a shoppable photo feed for Prime members where you posted products you loved and others tapped "smiles" instead of likes. The problem was in the premise. It was built around buying, not around people, so it felt transactional and bland, and it lacked the reasons anyone opens Instagram. About 10,000 Amazon customers used it on day one. Two years later, in 2019, Amazon quietly shut it down.

Product discontinuation Discontinued High
Company
Amazon
Started
2017
Ended
2019
Amazon customers who used it in its first 24 hours
~10,000
Collapse speed
Rapid
Preventability
High
Lesson transfer
Industry-wide
Last reviewed
2026-07-24

Narrative

The story

The ambition

Amazon wanted the part of shopping that happens on Instagram to happen on Amazon. Launched on July 18, 2017, Spark was a shoppable feed of stories and photos aimed at Prime members, an image-heavy stream of products people posted and recommended, each tagged so you could buy it in a click. Users reacted with "smiles" rather than likes, earned "Enthusiast" badges for contributing, and picked interest categories to shape their feed. The strategic goal was clear: capture the inspiration-to-purchase moment, the browsing that usually happened on Instagram or Pinterest, and keep it inside Amazon where a click meant a sale.

The rise

On paper Spark had an advantage its rivals lacked: less friction between wanting something and owning it. Where an Instagram photo left you to hunt for the product and hope it was in stock, a Spark post let you buy the item directly. Amazon had the catalog, the Prime base, and the checkout, and it was betting that removing the last steps between desire and purchase would make social shopping finally work.

The cracks

It never felt like a place anyone wanted to spend time. Spark was, by most accounts, bland: it focused on the shopping and the sale rather than on people, so everything felt transactional, and it lacked the things that make Instagram compelling, following friends, Stories, the sense that posting made you look good. As one analysis put it before launch, the qualities that make someone a top Amazon reviewer are not the qualities that make someone worth following in pictures, and Amazon was simply not "cool" for that kind of browsing. The numbers were tiny from the start: only around 10,000 Amazon customers used Spark in its first 24 hours, a rounding error for a company with hundreds of millions of shoppers.

The collapse

Spark faded rather than crashed. It never built the critical mass a social feed needs, and it was squeezed as Instagram itself added shopping features in late 2018 and in-app checkout in early 2019, doing social commerce from the social side that Amazon was trying to do from the commerce side. When Chee Chew, the Amazon vice president who had championed Spark, left the company for Twilio in early 2019, the product lost its internal advocate, and Amazon evidently decided it was not worth continuing without him. In June 2019, after about two years, Spark quietly disappeared, its web address redirecting to a new discovery effort, #FoundItOnAmazon.

The aftermath

Amazon did not abandon the ambition, only this version of it. It folded Spark's lessons, and those of a tool called Interesting Finds, into #FoundItOnAmazon, a narrower discovery feed for fashion and home décor, and years later would try again with a TikTok-style feed called Inspire. Spark stands as the first of Amazon's repeated, unconvincing attempts to bolt a social network onto a store.

The lessons

Social built around transactions is not social. Spark inverted what makes these feeds work: people come to Instagram for people, and shopping happens as a side effect, whereas Spark led with the shopping and hoped the community would follow, which left it feeling like a catalog with comments. The one-click advantage Amazon prized turned out to solve a problem that was not the bottleneck; the bottleneck was giving anyone a reason to open the app at all. There is also a lesson about internal fragility: a product that depends on a single executive's enthusiasm is one resignation away from cancellation, and Spark's quiet death the year its champion left is a reminder that thin adoption plus a lost sponsor is usually fatal. Whether Spark was a doomed concept or merely an orphaned one, the underlying miss was the same, assuming a store could become a place people gather.

Causal timeline

Failure Anatomy

  1. 2017-07-18

    Amazon's Instagram

    Launched July 18, 2017, Spark was a shoppable feed of photos and stories for Prime members, with "smiles" instead of likes, built to keep inspiration-to-purchase inside Amazon. [1]

  2. 2017

    Transactional and bland

    Spark focused on shopping over people, lacked following friends and Stories, and drew only about 10,000 users in its first 24 hours. [2] [4] [5]

    No real demandInformation failure
  3. 2018

    Squeezed by Instagram

    Instagram added shopping features in late 2018 and in-app checkout in early 2019, doing social commerce from the social side. [6]

    Stronger competitor
  4. 2019

    Champion departs

    The Amazon VP who championed Spark left for Twilio in early 2019, and Amazon apparently judged the product not worth continuing without him. [7]

    Leadership failure
  5. 2019-06

    Quiet shutdown

    Around June 2019, after about two years, Spark shut down and its web address redirected to [8]

    No real demand

Structured analysis

What Went Wrong

Root causes

Transactional, not social. Spark led with shopping rather than people, lacked following friends and Stories, and felt bland and transactional, so few had a reason to open it. [2] [3]

Amazon was not "cool" for this. The qualities that make a top Amazon reviewer are not the qualities that make someone worth following in pictures, and Amazon lacked the aspirational pull social feeds need. [4]

Contributing factors

Instagram added shopping. Instagram introduced shopping features in late 2018 and in-app checkout in early 2019, doing social commerce from the social side Amazon could not match. [6]

Lost its champion. The Amazon vice president who championed Spark left the company in early 2019, removing the product's internal advocate. [7]

Immediate trigger

Quiet shutdown. Never reaching critical mass and having lost its champion, Spark was shut down around June 2019, its web address redirecting to [8]

Visible symptoms

Almost no users. Only about 10,000 Amazon customers used Spark in its first 24 hours, negligible for a company with hundreds of millions of shoppers. [5]

Warning signs

Doubts before launch. Analysts questioned before launch whether Amazon could build the critical mass a social feed needs, given it was not seen as "cool" for that kind of browsing. [4]

Affected groups

Customers

Contested

Disputed points

Interpretations where credible accounts genuinely differ, presented as disputes, not settled facts.

Did Spark fail because the concept was doomed or because it lost its champion? One reading centers the premise, that a shopping-first feed built around transactions rather than people could never be socially compelling. Another centers organization, noting Amazon shut Spark down the same year the executive who drove it left, treating it as an orphaned project. Both point to the same underlying weakness, thin adoption with no one left to defend it. [3] [7]

Mixed

Evidence

Claims & sources

Every numbered marker in the analysis links to the claim it rests on, and each claim to its sources.

  1. [1]

    Amazon launched Spark on July 18, 2017 as a shoppable Instagram-style feed of photos and stories for Prime members, with product-tagged posts and "smiles" instead of likes, to keep social-commerce discovery inside Amazon.

    High Fact Amazon launches Spark, a shoppable feed of stories and photos aimed at Prime members Amazon Spark, the retailer's two-year-old Instagram competitor, has shut down
  2. [2]

    Spark focused on shopping rather than people, lacked following friends and a Stories feature, and felt bland and transactional.

    High Fact Amazon Spark, the retailer's two-year-old Instagram competitor, has shut down
  3. [3]

    Spark led with the shopping and the sale rather than with community, so it lacked the reasons people open Instagram.

    High Reported explanation Amazon Spark, the retailer's two-year-old Instagram competitor, has shut down Amazon Spark, A Social Network for Product Discovery
  4. [4]

    Analysts noted before launch that the qualities of a top Amazon reviewer are not those that make someone worth following in pictures, and doubted Amazon could build the critical mass a social feed needs.

    High Analysis Amazon Spark, A Social Network for Product Discovery
  5. [5]

    Only about 10,000 Amazon customers used Spark in its first 24 hours.

    High Fact Amazon Spark, the retailer's two-year-old Instagram competitor, has shut down
  6. [6]

    Instagram added shopping features in late 2018 and in-app checkout in early 2019, competing in social commerce from the social side.

    High Fact Amazon Spark, the retailer's two-year-old Instagram competitor, has shut down
  7. [7]

    Chee Chew, the Amazon VP who championed Spark, left for Twilio in early 2019, and Amazon judged the product not worth continuing without him.

    High Fact Amazon Spark, the retailer's two-year-old Instagram competitor, has shut down Amazon closes down its shopping-focused Instagram rival Spark
  8. [8]

    Around June 2019, after about two years, Amazon shut down Spark, and amazon.com/spark redirected to a narrower discovery effort,

    High Fact Amazon Spark, the retailer's two-year-old Instagram competitor, has shut down Amazon closes down its shopping-focused Instagram rival Spark

Sources