- Blippar has gone through several pivots since it was founded as an augmented reality startup in 2010.
- Sources tell Business Insider the company has burned through millions in the process.
- Blippar has raised £79 million in investment so far but will soon need more cash, according to sources.
- Senior staff say only a fraction of the Blippar Group user numbers are actively using the Blippar app, and that the company has little recurring revenue.
Blippar, one of the best-funded tech startups in the UK, is burning through cash, has low user engagement numbers, and is struggling to nail down its business model, former employees have told Business Insider.
Speaking anonymously, six former staff members who worked on different areas of the company's business, said the story Blippar has promoted publicly about its augmented reality (AR) and image recognition technology misrepresents the company's true reach.
According to an interview founder Ambarish Mitra gave to The Financial Times in 2015, Blippar is a unicorn — worth more than $1 billion to its investors. Mitra once told Business Insider that "what I'm trying to build is bigger than the internet itself."
The Blippar Group touts "65 million users of our technology" publicly. But internally, staff have told Business Insider that the company's main product, the Blippar app, represents only a small fraction of these users. With regard to these users, Blippar told Business Insider: "We don't disclose monthly active users."
The former staff estimated the company was burning around $3 million (£2.4 million) a month. Back-of-the-envelope maths would give the company a year to raise another round, although the company denies this is the case.
A spokeswoman said in a statement to Business Insider: "As a private company, we do not comment on our monthly expenditure figures and our cash flow. However, we can confirm that these claims are not correct."
The company also said it was focused on a "long-term vision" for its technology, and it would take time to get this right. "As with any company breaking new ground, we know that what we’re trying to do will take time."
"But with the recent launch of the world’s first true visual discovery browser (eventually allowing users to search through sight rather than words); a suite of tools to empower anyone to create their own AR experiences; and with the upcoming launch of a range of consumer-facing features, we are excited about the direction that we are taking."
From Delhi slum to London unicorn
TwitterBlippar was cofounded in 2010 by four people — CEO Ambarish "Rish" Mitra, CTO Omar Tayeb, CMO Jess Butcher, and chief creative officer Steve Spencer.
Of the four, 40-year-old Mitra has the highest profile due to his frequent media appearances.
At the age of 16 he ran away from his middle-class Indian home and lived on his own in a shed in a Delhi slum, he told Fast Company. He won a startup pitch contest he read about in a newspaper and used the $10,000 prize to create a web portal for women called WomenInfoline. He later took the company public in an IPO, according to the BBC. And he had 125 employees at WomenInfoOnline's height before he moved to London.
His life story sounds like a movie script and, according to The Financial Times, it has been edited like one over the years. The newspaper recently found that WomenInfoOnline had not been taken public. Mitra disputes the Financial Times report as “inaccurate” and explained to his staff in an internal memo sent in March, and later leaked to Business Insider, that he had only handed over the portal to a different company which was already public. An Economic Times article quoted Mitra as saying he once studied at LSE, but the school has no record of his attendance. Mitra has since said he was misquoted and acknowledged that he did not attend LSE.
No matter. The past is the past.
What worries Blippar's investors and employees now is the future.
Sources said Blippar is in a tough financial position
Blippar began as an augmented reality app for advertisers but pivoted in March last year to become an artificially intelligent visual search engine that lets users identify whatever they point their phone cameras at. It has received £79 million in investment (about $99 million in total) and at one point was valued at $1.5 billion (about £1.1 billion) in a putative takeover deal. Mitra said he rejected the offer, from a US company in 2013, because "We believe we’re going to be one of the biggest businesses in the world.”
The company is still best-known for its early AR offering, which let people scan (or "blipp") product packaging to bring up extra content on-screen. For brands, the idea was a way to make their product packaging interactive. For users, the sell was that you could get fun content like games from brands you liked.
In Blippar's early days, brands were genuinely excited by the technology and willing to spend money on it, two agency sources said. But their enthusiasm has waned as user-numbers failed to materialise.
The change in strategy from an AR app for brands to visual search recognition has put Blippar in a difficult financial situation, multiple sources said.
Mitra has said that he wants the all-new Blippar to be able to recognise "trillions" of objects, but this will take time and money. "As they've shifted into computer vision and a visual browser, there's been a lot of investment in engineering out in California, but it's a massive [computing] challenge," one source said.
This has meant a shift towards hiring engineers and a reduction in sales staff.
In the last year, Blippar shut international offices in Japan, Turkey, and India. It isn't clear how many staff were affected by the closures. Blippar also trimmed down its London team. Glen Drury was hired as chief commercial officer at the beginning of last year, but left less than six months later, according to his LinkedIn profile.
The company hired British diplomat Danny Lopez as COO in May. Lopez was previously Consul General in New York, where he was responsible for promoting the UK tech scene in the US. Lopez oversees global development at Blippar and is based in London.
The company hired engineer Xuejun Wang in 2015 to lead its efforts in computer vision and has employed more engineers in its San Francisco office.
Blippar confirmed the office closures, but said it had 70 international "partners" who were reselling its technology to advertisers. The company said in its statement: "The consolidation of our international offices will allow us to focus our resources on the geographies that offer the greatest potential for long-term growth."
Blippar is "burning" through cash, according to records and sources
Two of the former staffers said they left because Blippar was "burning" through money, estimating spend of around $3 million a month (£2.4 million).
One said that company had spent its 2015 round of funding in less than a year. That round was reported as $45 million (£36 million). Companies House filings show two lots of equity, one at £11 million ($12 million) in 2013, and another at $25 million (£20 million) in 2014, for a total of £31 million ($37 million). At the time, Blippar told The Guardian that the funding was being released in stages.
"They could never meet investor expectations," the person said.
Blippar told Business Insider: "All allocated funding initiated by investors has been received."
"In terms of expectations, the team at Blippar, and our investors, are focused on a long-term vision to bridge the physical and digital worlds, helping people to understand the world and the opportunities around them in an instant through the power of sight," the company added.
The second source said Blippar would, at its current rate, burn through its 2016 round of funding ($54 million / £43 million from Khazanah Nasional Berhad), towards the end of 2017.
The two sources estimated a burn rate of around $3 million (£2.4 million) a month.
Blippar's own figures show that the company was burning between £1 million to £2 million a month in 2016.
In the 16 months to March 2016, Blippar made £8.5 million in revenue, on a loss of £26 million. On average, the company was losing £1.6 million a month over this period from its underlying business.
The company's filings also show it spent £23 million in cash over the period. Over the 16 months, this averages out as £1.43 million a month in negative cashflow.
But Blippar's costs are likely to have gone up after its pivot, meaning our sources' estimates of £2.4 million cash burn a month are reasonable.
Here's a summary of Blippar's finances:
- Blippar has raised £79 million to date from Qualcomm Ventures, property developer Nick Candy, and Khazanah Nasional Berhad — the Malaysian government's investment arm.
- In the 16 months to March 2016, Blippar made £8.5 million in revenue, up from £4.5 million in 2014.
- Losses were £26 million, up from £5 million in 2014.
- Staffing costs were up six-fold to £12 million, from £2.2 million.
- The number of tech staffers has risen from 26 to 60.
Blippar's active user numbers are low, sources said
Blippar claims to have 65 million users "across the Blippar Group." According to a spokeswoman: "The Blippar group has the Blippar app, Layar app, and several white label apps that use our SDK. The combined number is over 65 million users of our technology. We don't disclose monthly active users." Blippar used the 65 million figure when promoting a campaign with magazine publisher Hearst last November.
According to the former employees, this doesn't actually mean Blippar has 65 million monthly active users, or even that all those downloads are for the Blippar app. The number of active users of the Blippar app is closer to 500,000, sources tell us.
"We don't share our active user numbers because the figures vary month to month, depending on the number and scope of the brand campaigns. At present, a large proportion of our audience belongs to the large brands for whom we develop our AR campaigns," the company said.
The sources — and public records — said the bulk of its users come from Blippar’s 2014 acquisition of Dutch AR app Layar.
Shona Ghosh/Business Insider
Layar is still a standalone app and, prior to its acquisition, was pre-installed on smartphones like the Samsung Galaxy Smartphone S. You can use it to scan QR codes or Layar logos to bring up extra content on-screen. When it announced the acquisition, Blippar said the deal gave it "one of the biggest data footprints for consumer AR use around the world."
Before the acquisition boosted its numbers, one source with knowledge of the matter said Blippar's monthly users were actually falling. This person claimed that Blippar’s monthly users were 518,000 in 2013, and had fallen to 504,000 in 2015.
"Blippar uses Layar's numbers to great effect," the person said. "They will say 'The Blippar Group', but the majority are actually Layar users. Maybe they're using the Blippar [software developer kit] now, but they are not users in a regular sense."
Another said the commercial team was instructed to give out that 65 million number, but would give out the real figure if directly questioned by ad agencies.
"It was certainly a concern for agencies. That was one of the biggest pieces of feedback, agencies tend to ask the right questions and they are a bit more keyed into what's going on," one former staffer said. "It was a roadblock in terms of being able to sell in projects. They were very wary of the real number of Blippar users versus the reality."
Ad agencies said Blippar started well ... but didn't grow
One agency source who had run a campaign with Blippar told Business Insider that engagement rates were "shockingly bad," though they wouldn't give numbers.
One of the former Blippar employees said engagement rates would differ depending on the campaign, but that they weren't consistent enough to keep driving revenue.
Another media agency CEO said that at one stage, Blippar had initially been popular with advertisers who wanted to try innovative tech, particularly when the app first came out.
"Advertisers were prepared to allocate innovation budget to Blippar, and it was around the time when lots of brands were toying around with activating packaging through QR codes," the person said. "It's quite a big media channel, when you think about packaging."
Cadbury used Blippar in 2012. The chocolate brand let people scan wrappers with Blippar to bring up a 30-second game on their phones. A senior Cadbury marketer said at the time that she "loved" Blippar, and that she was "blown away" by the technology.
Why would you download an app to have brands push advertising content at you?
But persuading users to download the app was harder, and that's when Blippar's problems began. As several sources said to Business Insider, why would you download a standalone app to have different brands push advertising content at you? If you scanned a can of Coca-Cola once, it isn't clear why you would ever reopen the app to scan another.
One former employee said: "Blippar's technology — the idea of what they're doing is interesting, but they don't have the groundswell to force users to take up the app and use it."
External data tells a similar story. According to the analytics startup SimilarWeb, people who download the dedicated Blippar app don't use it very often. Less than 0.4% of UK Android users return to Blippar 20 days after installing it, SimilarWeb's charts show.
That compares to an average of 4.8% across other "entertainment" apps. That doesn't sound so bad, until you compare it with a top performing app like Snapchat, which is much newer than Blippar and also offers AR. Snapchat's engagement rate is around 18% in the UK.
The figure for Blippar is 0.5% in the US, compared with 22% for Snapchat, and 6% on average.
Blippar said it had "consistently delivered for brands." The company said users spend an average of 85 seconds engaged with a Blippar campaign, and said clients often repeated business across different products or countries.
The company added: "In terms of the use of Blippar outside of campaigns where users can see the Blippar logo, we are continuously working on improving the recognition and content experiences. We know that our vision is ambitious and will take time to accomplish."
"The company is unfortunately quite happy to pivot, but it's almost like spinning rather than pivoting"
Blippar's technology has gone through several major changes which, the sources said, resulted in a "mixed business model." One said: "The company is unfortunately quite happy to pivot, but it's almost like spinning rather than pivoting."
This is how Blippar has changed its core technology over the years, according to the former employees and publicly available records.
In 2012, two years after Blippar was founded, a Qualcomm press release stated that the company based its AR technology on a Qualcomm-owned AR service called Vuforia.
In 2014, Blippar bought Layar and began describing Vuforia as "a competitor", according to a TechCrunch report quoting CEO Ambarish Mitra. "They started to develop their own capabilities," one source explained.
At the end of last year, Blippar had pivoted from AR campaigns to identifying objects.
If you open the current version of the Blippar app it will open straight into the camera. If you point your phone at different everyday objects, Blippar can theoretically identify them and offer up extra information. For example, if you "blip" a real-life mug of tea, Blippar will bring up a definition of tea, recipes, and suggestions for a nearby cafe on-screen.
Shona Ghosh/Business Insider
It also promised facial recognition in December 2016 but the technology has not yet rolled out to users. If you open the app, there's a holding page that promises facial recognition.
Shona Ghosh/Business Insider
Blippar acknowledged that its strategy has changed over the years. It said: "At only six years old and as a technology company, it is natural and important that Blippar adapts and evolves its strategy over time ... As such, we have built our capabilities in the areas of AI, computer vision and deep learning through a team of world-class engineers, split across our UK, US and India offices, which has over 175 patents to its name."
The company said it successfully launched facial recognition for public figures in December, adding: "We have some exciting developments on the facial recognition piece but we can’t share more before the launch."
Blippar's image recognition is "like a toddler doing a Mastermind challenge"
Blippar's financials suggest the company has limited time to expand its object recognition technology. Experts told Business Insider that the current version of the app is limited in functionality.
One computer vision expert, UCL computer science professor Peter Bentley, described the app during an interview with Business Insider as "a toddler doing a 'Mastermind' challenge."
"It's an amusing toy to play with, but it's not ready to be a serious application without a lot more data," he said. "This is the same reason why Google has been driving its self-driving cars for several years. Wherever the cars go, they are generating more data to learn what roads look like." He added: "[Blippar's] datasets are large, but it doesn't have enough examples of different things to understand what it is looking at."
Stephen Hicks, research fellow at the University of Oxford, praised Blippar for bringing AR to mobile but also said the firm would be limited by its datasets, and how images are labelled.
"The weakness is that it's tested on datasets of still images that can be from anywhere ... most startups don't have the ability to find real-world images from the user's perspective, so [they] have to revert to libraries," he told Business Insider. "Then they are at the mercy of whoever labelled them to begin with."
When Business Insider experimented with Blippar’s object recognition, the app identified a foot as a hand, a dining room table as a poncho, and a chair as a cradle. It did also correctly identify flooring, a record player, and a coffee table.
Shona Ghosh/Business Insider
Blippar said its visual recognition capabilities could be compared to "those of a child with 33,000 words to describe the world in a generic level."
The spokeswoman said: "In some areas, Blippar’s app has a far more nuanced level of understanding and an exceptional rate of accuracy. However, it takes time to index the physical world - a process that is now being made much quicker by the use of AI. Over time the recognition will become more accurate in all areas."
"Revenue doesn’t match and follow their investment, so they constantly have to sell the big dream again"
Where does all of this leave Blippar in 2017?
A Bloomberg report, and sources speaking to Business Insider, said the company is trying to raise money again. Blippar wouldn't comment on the report.
One source said: "People put money [into Blippar] because they believe what Rish says. Big dreams cost a lot of money to make ... But the growth of revenue doesn’t match and follow their investment, so they constantly have to sell the big dream again."
Outside of the company, competitors have closed in.
Shona Ghosh/Business Insider
When Business Insider spoke to media agency sources last month, one said: "One way for Blippar to become a serious scaleable option is if it's integrated into the camera on a phone, and you don't have to open a separate app — you just point your phone at something. Then it's a B2B player where Blippar sells its tech to a mobile phone manufacturer. That's the only way."
The person's prediction was uncanny. Shortly afterwards, Samsung launched the Galaxy S8 with object recognition technology built into the camera, called Bixby Vision, with visual search powered by Pinterest.
It's not clear how Blippar could ever hope to compete with that.
There are 48 reviews of Blippar left on the employer review site Glassdoor. These vary in sentiment, but one theme emerges consistently — that the startup needs to decide on a strategy.
One anonymous reviewer wrote in December: "There isn't a clear vision, teams work along each other without proper alignment, there's no overarching engineering culture."
Another wrote in August: "Continuous innovation is great. But, strategy can't change every 6 months. Decide on what you want to identify as the core competency, and develop a realised strategy based on that."
The same user added, perhaps presciently, "If not the end is near."