American Express quietly acquired UK fintech startup Cake for $13.3M

Cake Technologies, the U.K. fintech startup that wanted to make it more convenient to pay your restaurant or bar bill, has been acquired by American Express — as the credit card behemoth plans to beef up its payment options for Amex members.

According to sources the deal quietly completed in October last year for a final price of $13.3 million (approx. £10.1m). However, due to an eleventh-hour preferential debt round and after fees, only some shareholders made a profit. I also understand from one source that Cake had raised a total of £4.5 million in equity and £1.4 million in debt. Part of the equity funding was a £1 million crowdfunding round on Crowdcube in 2015.

Confirming the acquisition, American Express gave TechCrunch the following statement:

Last year American Express acquired Cake Technologies. This year, we will be on-boarding Cake and their technologies to collaborate on ways to provide our Card Members with enhanced service and value in the dining space, which is an area many of our Card Members are passionate about.

A spokesperson for American Express declined to comment on the exact financial terms of the deal, but said that it was a “good outcome for Cake employees, previous investors and American Express”. They did confirm, however, that Cake employees are now employees of American Express.

This includes Cake founders Charlotte Kohlmann and Michelle Songy, who hold the positions of Vice President Global Dining Platform Solutions at American Express, and Director Global Platform Dining Solutions at American Express, respectively.

“We are excited to have Cake on board with us and look forward to collaborating on bringing our Card Members exciting new capabilities in the dining space soon,” adds the American Express spokesperson.

The back story to Cake’s eventual exit makes for interesting reading. According to a source with knowledge of the startup’s path to a sale, who spoke to TechCrunch on the condition of anonymity, it was very close to raising a £5 million Series A in the fall of 2016 before the company’s founders walked away for “ethical reasons,” although the source declined to diverge what these were. This then left Cake in a precarious situation financially as the company could not find another VC to step in quickly enough before running out of cash.

In the holidays/early 2017, the board of Cake put together a rescue round that was structured in the form of debt and designed to give the startup more runway to try to achieve a trade sale. All existing shareholders were given the chance to participate on a pro rata basis, although some declined due to the substantial risk of doubling down.

The loan was also structured so that, should the company get acquired, these eleventh hour investors would get a multiple preferential return. This, I’m told, explains why some investors made money from the exit, while others, including some Crowdcube backers, lost money, even possibly after factoring in EIS tax breaks.

In May 2017, American Express first made an offer to acquire Cake. The startup passed due diligence in late June, but American Express pulled the deal in mid-July for unknown reasons. Determined to get the sale back on track, Cake co-founder Kohlmann flew to New York unannounced and the deal eventually closed in October.

“Despite the complications and lengthy process, Amex did a really good deal here,” says my source. “It is clear that Cake is now a very important part of their digital strategy and the purchase price looks like good value in that context. Cake’s user experience will be a benefit to users of the Amex app once fully integrated and Cake’s basket level POS integrations will give Amex better insight into exactly what products their customers are buying rather than just where they go and how much they spend”.

Tumblr confirms 84 accounts linked to Kremlin trolls

Tumblr has confirmed that Kremlin trolls were active on its platform during the 2016 US presidential elections.

In a blog post today the social platform writes that it is “taking steps to protect against future interference in our political conversation by state-sponsored propaganda campaigns”.

The company has also started emailing users who interacted with 84 accounts it now says it has linked to the Russian trollfarm, the Internet Research Agency (IRA).

In the blog post it says it identified the accounts last fall — and “notified law enforcement, terminated the accounts, and deleted their original posts”.

“Behind the scenes, we worked with the Department of Justice, and the information we provided helped indict 13 people who worked for the IRA,” it adds.

In an email sent to a user, which was passed to TechCrunch to review, the company informs the individual they “either followed one of [11] accounts linked to the IRA, or liked or reblogged one of their posts”.

“As part of our commitment to transparency, we want you to know that we uncovered and terminated 84 accounts linked to Internet Research Agency or IRA (a group closely tied to the the Russian government) posing as members of the Tumblr community,” the email begins.

“The IRA engages in electronic disinformation and propaganda campaigns around the world using phony social media accounts. When we uncovered these accounts, we notified law enforcement, terminated the accounts, and deleted their original posts.”

Last month Buzzfeed News — working with researcher, Jonathan Albright, from the Tow Center for Digital Journalism at Columbia University — claimed to have unearthed substantial Kremlin troll activity on Tumblr’s meme-laden platform — identifying what they dubbed as “a powerful, largely unrevealed network of Russian trolls focused on black issues and activism” which they said dated back to early 2015.

The trolls were reported to be using Tumblr to push anti-Clinton messages, including by actively promoting Democrat rival Bernie Sanders.

Decrying racial injustice and police violence in the US was another theme of the Russian-linked content.

Since then The Daily Beast has also reported on leaked data from the IRA which also implied agents at the trollfarm had used Tumblr — and also Reddit — to spread political propaganda to target the 2016 US election.

Those IRA leaks suggested the IRA had created at least 21 Tumblr accounts — and included names replete with slang terms, including some accounts listed in the user email we’ve reviewed.

Tumblr, which is owned by TechCrunch’s parent company Oath, did not respond to an email we sent to their press office last month asking about possible Kremlin activity on its platform.

In today’s public post, the company writes: “As far as we can tell, the IRA-linked accounts were only focused on spreading disinformation in the U.S., and they only posted organic content. We didn’t find any indication that they ran ads.”

As well as emailing affected users, Tumblr says it will be keeping a public record of usernames linked to the IRA or “other state-sponsored disinformation campaigns”.

The full list of 84 Kremlin accounts on its public page is as follows:

It also suggests users step in and “correct the record” when they see others spreading misinformation, regardless of whether they believe it’s being done intentionally or not.

Concluding its email to the user who had unwittingly engaged with 11 of the identified IRA accounts, Tumblr adds: “We deleted the accounts but decided to leave up any reblog chains so that you can curate your own Tumblr to reflect your own personal views and perspectives.

“Democracy requires transparency and an informed electorate and we take our disclosure responsibility very seriously. We’ll be aggressively watching for disinformation campaigns in the future, take the appropriate action, and make sure you know about it.”

Asked how he feels to learn Kremlin trolls had unknowingly infiltrated his Tumblr feeds, the user told us: “It’s unsettling, although maybe not surprising, that we legitimize and signal boost bad actors on social platforms by ‘liking’ or reposting content that doesn’t appear to have any political agenda at first glance.”

Telegram chalks up 200M MAUs for its messaging app

Another usage milestone for messaging platform Telegram: It’s announced passing 200M monthly active users “within the last 30 days”.

The platform passed 100M MAUs back in February 2016, when it held a lavish party at the Mobile World Congress tradeshow in Barcelona to celebrate the metric. At the time it said it was adding 350k new users daily and that there were 15 billion messages generated daily.

Since then Telegram has kept its powder fairly dry on the usage metrics front — presumably waiting to be able to announce 200M.

Its blog post is not revealing of any other details about usage. Rather founder Pavel Durov uses the space to give thanks to Telegram users for getting the company to the milestone, and takes a sideswipe at other “popular apps” which he says — unlike Telegram — monetize via advertising and/or pass data on to third parties.

Safe to say, it doesn’t take much imagination to figure out who he might be thinking of

“Since the day we launched in August 2013 we haven’t disclosed a single byte of our users’ private data to third parties,” he writes (emphasis his). “We operate this way because we don’t regard Telegram as an organization or an app. For us, Telegram is an idea; it is the idea that everyone on this planet has a right to be free.”

We’ve reached out to Durov to see if he’ll give up any more Telegram usage tidbits and will update this post if so.

While he writes confidently now that “Telegram doesn’t… do deals with marketers, data miners or government agencies”, it’s not clear how much longer he’ll be able to stand up that claim — given the legal pressure being applied, for example, in Russia to hand over encryption keys or face being blocked. Telegram has also faced restrictions in Iran.

It told Bloomberg it plans to appeal the Russian ruling in a process that may last into the summer, according to company lawyer, Ramil Akhmetgaliev.

Durov also tweeted that: “Threats to block Telegram unless it gives up private data of its users won’t bear fruit. Telegram will stand for freedom and privacy.”

Meet the startups that pitched at EF’s 9th Demo Day in London

Entrepreneur First (EF), the company builder and “talent first” investor, held its ninth London Demo Day this afternoon. Once again, the pitches took place in front of a packed crowd at King’s Place in London’s King Cross area, seeing 19 startups pitch their wares to investors, press and other actors in the European tech scene.

EF stands out from the plethora of demo days that the U.K. capital city hosts because of the way the investor backs individuals “pre-team, pre-idea” — meaning that the companies pitching only came into existence over the last 6 months and perhaps may never have done so without the founders entering the programme.

As is now a tradition, prior to the pitches, EF co-founder Matt Clifford took the chance to announce some EF news of its own. Already operating in London, Singapore and Berlin, the company builder — which last year picked up backing in a $12.4 million round led by Silicon Valley’s Greylock Partners — is expanding to Hong Kong to double down on its Asia ambitions, kicking off with a local programme in July. Heading up EF’s Hong Kong office is former Airbnb and Google exec Lavina Tien — you can read my full report here.

The themes for EF’s ninth London Demo Day continued to reflect the company builder’s focus on recruiting the best technical and domain expert talent — both recent graduates and also people already working at tech companies — where they are encouraged to try their hand at entrepreneurship. The pitches spanned AI/machine learning, automation, healthtech, legal tech, financial services, new interfaces and input devices. And, being that this is 2018, the blockchain and cryptocurrency, although, thankfully, there wasn’t an ICO in sight.

Low on sleep and therefore particularly prone to pitch-lash, my picks this time around were Limbic, which wants to bring emotional intelligence to software (and therefore hardware) by measuring changes to a part of your brain called the ‘limbic system’ via your heartbeat; nPlan, which wants to use machine learning to enable major construction work schedules to be more accurate and run on or ahead of time; and Inoviv, which is developing technology based on unique insights into proteomics to help match patients with the right drugs.

The full list of presenting teams (in their own words)

Papercup translates the voice track on videos so every creator can expand their reach to seven billion people.
CreditMint is decentralising corporate lending.
nPlan accelerates the global construction industry.
Nivoda aggregates the world’s diamonds.
Headlight AI provides intelligent sensing and mapping software for harsh environments.
Octeract solves today’s unsolvable optimization problems.
Beneficiary maps the impact of the world’s philanthropic efforts and helps charities identify the most effective ways to improve lives.
Ginie AI turns term sheets into contracts.
Prime Factor Capital is a crypto asset manager.
Panopy is a performance management platform that removes workplace bias.
Mimica builds self-learning automations for digital work.
Inoviv matches patients with the right drugs.
PolyAI is democratising conversational AI to give machines a voice.
Plural AI is building a new kind of search engine: a knowledge engine.
ArrayStream Technologies helps fund providers launch next generation AI-powered mutual funds.
Limbic provides computers with vital emotional input.
Plumerai is making small machines intelligent.
TokenAnalyst is bringing transparency to the decentralized economy.
Massless is the future of three-dimensional immersive design.

Cambridge Analytica’s Nix recalled by fake news probe

Stock up on the popcorn — the currently suspended CEO of the firm at the center of a data handling and political ad-targeting storm currently embroiling Facebook, Cambridge Analytica, has been recalled by a UK parliamentary committee that’s running a probe into the impact of fake news because it’s unhappy with the quality of his prior answers.

The committee also says it has fresh questions for Alexander Nix in light of revelations that hit the headlines at the weekend about how a researcher’s app was used to gather personal information on about 270,000 Facebookers and 50 million of their friends, back in 2015 — data that was passed to CA in violation of Facebook’s policies.

Nix gave evidence to the DCMS committee on February 27, when he claimed: “We do not work with Facebook data, and we do not have Facebook data. We do use Facebook as a platform to advertise, as do all brands and most agencies, or all agencies, I should say. We use Facebook as a means to gather data. We roll out surveys on Facebook that the public can engage with if they elect to.”

That line is one of the claims the committee says it’s keen to press him on now. In a letter to Nix, it writes: “[T]here are a number of inconsistencies in your evidence to us of 27 February, notably your denial that your company received data from the Global Science Research company [aka the firm behind the survey app used by CA to harvest data on 50M Facebook users, according to The Observer].”

“We are also interested in asking you again about your claim that you “do not work with Facebook data, and […] do not have Facebook data,” it continues, warning: “Giving false statements to a Select Committee is a very serious matter.”

The self-styled ‘not a political consultancy’ but “technology-driven marketing firm” (and sometime “campaign consultancy and communication services” company) — which Nix also described in his last evidence session as “not a data miner… a data analytics company” — had its Facebook account suspended late last week for violating Facebook’s platform policies.

The UK’s data protection watchdog, the ICO, has also applied for a warrant to gain access to CA’s offices and servers — accusing the company of failing to hand over information the regulator had requested as part of a wider investigation it’s carrying out into the use of data analytics for political purposes.

CA is also now facing several legal challenges from Facebook users angry about how their data appears to have been misused.

We reached out to the company for comment on the DCMS recall. At the time of writing it had not responded.

Below are a few choice segments from Nix’s last evidence session in from of the committee — which we expect he will be asked to revisit should he agree to make a repeat appearance…

Q698       Rebecca Pow:… Could you expand a bit more on what those surveys are, what you are asking people and how you are gathering the data? Do you keep that data on surveys carried out on Facebook or does Facebook keep it?

Alexander Nix: I cannot speak to Facebook, but as far as I am aware the process works a bit like an opinion survey. If I want to find out how many people prefer red cars or yellow cars, I can post that question on Facebook and people can agree. They can opt in to answer a survey and they give their consent and they say, “I prefer a yellow car” and then we can collect that data. That is no different to running a telephone poll or a digital poll or a mail poll or any other form of poll. It is just a platform that allows you to engage with communities.

Q699       Rebecca Pow: Are they a big part of your data-gathering service?

Alexander Nix: When we work for brands, whether it is in the UK or in the US or elsewhere, we often feel the need to probe their customers and find out what they think about particular products or services. We might use Facebook as a means to engage with the general public to gather this data.

Q700       Simon Hart: Let me ask a very quick question on the Facebook survey opt-in option that you were describing. If you are asking somebody what kind of car they prefer and they opt in, does that facilitate access to other data that may be held by Facebook, which is irrelevant to car colour, or is it only the data you collect on car colour that is relevant?Nothing else that is part of the data held by Facebook would be available to you.

Alexander Nix: You are absolutely right—no other data. As far as I am aware, Facebook does not share any of its data. It is what is known as a walled garden, which keep its data—

Q701       Simon Hart: People are not in any way accidently giving you consent to access data other than that that you specifically asked for.

Alexander Nix: That is correct. People are not giving us consent and Facebook does not have a mechanism that allows third parties such as us to access its data on its customers.

Q702       Simon Hart: Even with its customers’ consent.

Alexander Nix: Even with its customers’ consent.

Chair: You said in your letter to me that, “Cambridge Analytica does not gather” data from Facebook.

Alexander Nix: From Facebook?

Chair: Yes.

Alexander Nix: That is correct.


Q718       Chair: The actual quote from the letter is: “On 8 February 2018 Mr Matheson implied that Cambridge Analytica ‘gathers data from users on Facebook.’ Cambridge Analytica does not gather such data.” But from what you said you do, do you not, through the surveys?

Alexander Nix: Yes, I think I can see what has happened here. What we were trying to say in our letter is that we do not gather Facebook data from Facebook users. We can use Facebook as an instrument to go out and run large-scale surveys of the users, but we do not gather Facebook data.

Q719       Chair: By that do you mean that you do not have access to data that is owned by Facebook?

Alexander Nix: Exactly.

Q720       Chair: You acquire data from Facebook users through them engaging with surveys and other things.

Alexander Nix: Exactly right.

Q721       Chair: Is your engagement, either directly or through any associate companies you may have, just through the placing of surveys or are there other tools or games or thingsthat are on Facebook that you use to gather data from Facebook users?

Alexander Nix: No, simply through surveys.


Q729       Chair: In that presentation I think there is a slide on data analytics where you describe that data is sourced from multiple sources and any marketing company will know that there are companies that specialise in data analytics to analyse consumer behaviour. I think on your chart you had logos of different companies. I think Experian was one and Nielsen was one. You had Facebook on there as well. Again, just to confirm on this, is that because you are highlighting the fact that you can gather data from Facebook?

Alexander Nix: Collect data through Facebook—that is exactly right, yes.

Q730       Chair: Does any of your data comes from Global Science Research company?

Alexander Nix: GSR?

Chair: Yes.

Alexander Nix: We had a relationship with GSR. They did some research for us back in 2014. That research proved to be fruitless and so the answer is no.

Q731       Chair: They have not supplied you with data or information?

Alexander Nix: No.

Q732       Chair: Your datasets are not based with information you have received from them?

Alexander Nix: No.

Chair: At all?

Alexander Nix: At all.

Entrepreneur First, the London-based company builder backed by Greylock, expands to Hong Kong

When Silicon Valley’s Greylock Partners led Entrepreneur First‘s $12.4 million funding round in September, Greylock’s Reid Hoffman said he could see the company builder expanding to “20 or 30 or 40 cities, maybe even 50“. Since then, EF has expanded to Berlin, in addition to existing programmes in London and Singapore, and today the so-called ‘talent first’ investor is adding Hong Kong to the list.

Heading up EF’s Hong Kong office is former Airbnb and Google exec Lavina Tien, while the Hong Kong programme, which kicks off in July, will copy the Berlin format, meaning that it will run for 3 months per cohort, not 6 months as in London and Singapore. In addition, teams formed at EF Hong Kong will be eligible to participate in its Singapore demo day.

This is part of a new EF format that aims to make the company builder’s secret sauce, which sees it recruit founders ‘pre-team, pre-idea,’ a lot more scalable. So far, EF co-founder Matt Clifford tells me, it’s working out well.

He says the Berlin program was able to set up and recruit its first cohort in 9 weeks compared to the 9 months it took to get fully operational in Singapore, sounding extremely bullish about the future potential for more expansion.

That’s because the new shorter formula is designed to let EF focus locally on the part most unique to the organisation — persuading the best technical and domain talent to try their hand at entrepreneurship and in turn matching them with a complementary co-founder so that they can form a startup that might otherwise never exist.

Clifford also says this is about doubling down on EF’s Asia ambitions. He notes that, similar to other EF outposts, Hong Kong is a burgeoning but perhaps latent tech ecosystem with good education — such as Hong Kong University for Science and Technology, the University of Hong Kong, and the Chinese University of Hong Kong — and access to capital that is beginning to turn its attention locally rather than simply investing abroad.

Adds EF co-founder Alice Bentinck: “We believe that there are a handful of exceptional technologists globally who have the skills and ambition to build the next generation of breakout technology companies. We know that we will find some of them in Hong Kong, just as we have in London, Singapore and Berlin”.

Meanwhile, Clifford won’t be drawn into where EF might expand next, although he doesn’t rule out adding a further programme this year. If I had to guess, I’d say Paris is a good bet, but in all honestly there are quite a number of cities that could tick the EF box.

Separately, I’m hearing that the company builder is raising a new investment fund so that it can continue the strategy of doing follow-on investments at seed and Series A into the most promising companies it helps build, across all of the locations it now operates. As always, watch this space.

UK data watchdog still waiting for warrant to raid Cambridge Analytica

The UK’s data watchdog, the Information Commission’s Office (ICO), has still not obtained a warrant to enter and search the servers of the London-based political consultancy, Cambridge Analytica — the company at the center of the data misuse scandal engulfing Facebook — three days on from beginning the process.

The earliest a warrant could now be obtained by the regulator is Friday.

In a statement today the ICO said: “A High Court judge has adjourned the ICO’s application for a warrant relating to Cambridge Analytica until Friday. The ICO will be in court to continue to pursue the warrant to obtain access to data and information to take forward our investigation.”

The information commissioner, Elizabeth Denham, made it public on Tuesday that she was seeking a warrant to search CA’s servers after the company missed a Monday deadline to hand over information her office had requested. (Though in a statement on Tuesday CA claimed to “have been fully compliant and proactive in our conversations with the ICO”.)

She also instructed Facebook to withdraw its own investigators from CA’s offices, warning that their presence could compromise her investigation.

Unlike competition authorities, the ICO does not have legal powers to raid offices without a warrant. And former UK attorney general, Dominic Grieve, has argued the ICO’s legal powers are inadequate — telling the BBC on Tuesday that the Facebook-CA scandal highlighted a need for “greater powers and greater sanctions”.

Greater sanctions are at least incoming — under the EU’s GDPR regime which will apply from May 25, raising the maximum fine for the most serious data protection violations to up to 4% of a company’s global turnover (or €20M, whichever is greater).

But the fact that the data watchdog is forced to sit on its hands waiting to gain access to servers that the companies of interest to its investigation are in control of or able to access raises serious questions about the asymmetry between big data and regulation.

Earlier this month Denham told MPs on the DCMS committee that’s investigating fake news that her office would be pushing for increased transparency around data flows and disclosure rules for digital political advertising — suggesting a code of conduct is needed to regulate the use of social media in political campaigns, referendums and elections.

And while Facebook has claimed it was unaware that ~50M Facebook users’ data was passed to Cambridge Analytica for political targeting purposes, Facebook has itself long been actively encouraging politicians and political campaigns to make use of its tools — at a time when there was a complete lack of regulation for political ads on digital platforms.

Almost a year ago, in May 2017, the ICO announced a formal investigation into the use of data analytics for political purposes — including looking into complaints related to Cambridge Analytica’s use of data for ad targeting.

That investigation remains ongoing. And may well now be further delayed, given the developing nature of the story (and the ICO’s push for a warrant so it can conduct a full audit of CA’s servers).

Although, earlier this month before the latest Facebook-CA revelations hit the headlines, Denham told the committee she hoped to be able to publish the report by the end of May.

Asked by the DCMS committee whether the ICO has adequate powers to carry out its responsibilities Denham flagged a problematic gap in her “information notice powers” — noting that while the ICO can make a formal demand for information, organisations are not compelled to disclose the requested data (though they can be prosecuted for not doing so).

“Without the power to compel it is difficult to secure the desired outcome,” she told the committee. But she added she’s raised the issue with ministers and is hopeful the UK government will remedy this gap.

The founder of business banking startup Tide plans to step down as CEO

Changes are afoot at Tide, the U.K. fintech startup that offers banking services for small businesses. TechCrunch has learned that founder George Bevis is planning to step down as CEO, and that the nearly three-year old company is actively headhunting for his replacement.

It comes at a time when Tide — which counts 30,000 small business sign ups — is said to be entering ‘scale-up’ mode, with a headcount approaching 100 employees, and ambitions to expand internationally. Earlier this week the service saw a rebrand, including a new ‘vertical’ design for the Tide card and the slogan “Do Less Banking,” a reference to the startup’s mission to make the lives of small business owners easier.

The company also announced that it had got a regulatory upgrade and is now authorised as an electronic money institution by U.K. regulator the FCA. This gives Tide more direct access to banking infrastructure and means that over time it will be less reliant on third-party providers and can have more flexibility in how it serves customers, although it still hasn’t (yet?) chosen to apply for a fully fledged bank license.

Confirming that Tide is recruiting a new CEO, founder Bevis gave TechCrunch the following statement:

“I’m a small business-focussed guy who’s had the privilege of building an amazing company serving small businesses. Now our own business isn’t small any more it’s time for me to think about bringing in someone who knows at least as much about international scale-ups as I know about U.K. startups. Tide will stay focussed on saving small business owners time — in future all across the world. I’m looking forward to continuing to play a key role, both inside the business and on the board”.

I’m told that the decision to start recruiting for a new CEO was instigated by Bevis in discussion with the Tide board, who are fully supportive. The thinking from the Tide founder is that now is a good time to look for a CEO experienced in scaling a company as the early-stage founding job is materially complete, including developing the core Tide product and finding market fit.

Meanwhile, I understand that the new CEO will be tasked with executing Tide’s growth plans, which, along with international expansion, will include evolving the startup to a full SMB banking platform play that will see it continue to plug into providers of other bank related services for small business and further commercializing through revenue-share deals. The idea is that by creating a Tide ecosystem, the company “can scale far beyond the size of any single individual provider”.

To that end, Tide has secured over $16 million in funding to date from VCs including Creandum, LocalGlobe, Passion Capital and fintech specialist, Anthemis, as well as well-known angel investors including Errol Damelin (Wonga), Alex Chesterman (Zoopla/ZPG) and William Reeve (Lovefilm, Graze, and currently CEO of Goodlord).