Category: UNCATEGORIZED

“No bad conversations between companies and their customers is what we’re shooting for,” Kair Käsper tells me. He’s the Head of Growth of a relatively new startup called Klaus, which he founded together with old high school friend Martin Kõiva.

Most recently the pair were employees at Pipedrive, holding the roles of Director of Product Marketing and Global Head of Customer Support, respectively. Many years prior to that they shared a flat together and worked on a number of projects. One of those was an applicant tracking startup called Jobkitten “that didn’t really go anywhere”.

The latest Käsper and Kõiva venture, however, appears to already be on firmer footing. Described as a “conversation review and QA tool for support teams,” Klaus is designed to help companies improve the quality of customer service. Two years in the making but only launched formally 6 months ago, customers already include Automattic, Wistia, and Soundcloud. And today the Estonian startup is disclosing $1.9 million in seed funding led by Creandum, the first Baltic investment by the Swedish VC firm and the first from its new fund.

“The problem is that maintaining an even, high level of customer service quality is hard,” explains Käsper. “It becomes even harder if you have over 20,000 monthly conversations with customers and your support team is 100 people in 3 offices.

“As the head of customer support, you want everyone on your team to provide answers that meet with internal standards, regardless of how long they’ve been with the company or how seriously they take their job. You get very anxious in this situation, because you have no idea about what’s going on in those thousands of conversations. For you, no visibility means no control”.

He says that his and Kõiva’s first hand experience at Pipedrive taught them that the key to quality assurance is going through past interactions and then giving systematic feedback to agents. “Kind of like code review in engineering or the editorial process in writing,” he says. “Teams all over the world are discovering this now, but they almost always start with a manual process, managed in spreadsheets. They get stuck fast”.

To make this type of feedback loop more scalable, Klaus has created a purpose-built UI for giving internal feedback. Smartly, it also integrates with modern SaaS help desk solutions, such as Zendesk and Intercom.

“[The software also has] countless specialized features that allow you to focus on the actual feedback instead of managing a spreadsheet,” adds the Klaus Head of Growth. They include the ability to easily filter out conversations for review, rate them based on a customized score card and notify agents of received feedback through email or Slack.

Meanwhile, the young company makes money by charging a monthly or yearly subscription fee based on how many users are connected to its app. In other words, just like Pipedrive before it, another classic enterprise SaaS play out of Estonia.

Update: An earlier version of this article wrongly said that Kair Käsper is CEO of Klaus when his job title is actually Head of Growth.

Sony is doubling down on the world of startup investment. The Japanese tech giant announced a new fund that is aiming to raise 20 billion JPY (around $185 million) to invest in companies within “key high-growth industries.”

While Sony launched a fund in 2016, this new vehicle — which is called Innovation Growth Fund — has been set up with others. Firstly, it is being run jointly with Daiwa Capital Holdings — the VC arm of investment bank Daiwa Securities — and early LPs confirmed include Sumitomo Mitsui Banking Corporation, Osaka Shoko Shinkin Bank and Mitsubishi UFJ Lease & Finance Company Limited. Sony isn’t saying how much has been raised so far, but it isn’t the full target yet.

The previous fund made over 40 investments, Sony said, and now IGF is taking over with the goal of writing bigger checks than Sony typically manage by itself and paying closer attention to tech startups.

One longer-term goal is to help its portfolio companies develop into public firms, which is where Daiwa’s expertise of public listings comes into play. The fund, meanwhile, said it plans to open links with “renowned research institutions” and other tech companies to help its startups on their path — the latter certainly sounds like a SoftBank Vision Fund-style approach, albeit considerably less than its $100 billion ammunition.

“We believe that the integration of Sony’s insight of cutting-edge technologies and Daiwa Securities Group’s expertise in finance will lead to the creation of a new kind of venture capital business while providing the spark for new trends in the venture capital ecosystem,” said Yoshihisa Kaneko, executive managing director of Daiwa Securities, in a statement.

Cubyn, the Paris-based logistics startup that lets e-merchants outsource fulfilment and delivery logistics, has raised €12 million in new funding. The round is led by DN Capital, with participation from Partech Ventures, 360 Capital Partners, BNP Paribas Developpement, and the French investment bank BPI France.

The injection of capital is timed with the launch of “Cubyn Fulfillment,” as the company moves beyond pickup and delivery only. The new service is described as a fully-integrated “first mile” solution that covers the entire fulfilment process including keeping stock in Cubyn’s warehouses. It claims to be offered at a 30% lower price point than competitors.

“We want to make affordable world-class logistics accessible to every single e-merchant, whatever their size,” Cubyn co-founder and CEO Adrien Fernandez Baca tells TechCrunch. “Our typical customer is an e-merchant who sells across sales channels (marketplaces their own website). Size can go from 500 to 50,000 orders shipped per month”.

Launched in 2015, Cubyn says that in four years it has made over 2 million shipments. It also reckons that because its tech is “built from the ground up,” the startup is well positioned to tackle fulfilment more efficiently than legacy players.

“Most direct competitors are the traditional third party logistics players who missed the e-commerce revolution and lack technology intelligence,” says Baca. “We are 30% cheaper, with simpler multi-channel integrations and higher delivery quality. Less direct competitors are the fulfilment offer of marketplaces. They do offer a good logistics experience at a good price, but only for orders going through their marketplace”.

This, he argues, means there is a big gap in the market for a solution geared at multi-channel e-merchants. “We are marketplace agnostic and offer a seamless and high quality multi-channel logistics,” adds the Cubyn CEO.

Specifically, the way the new Cubyn Fulfilment product works is as follows: An e-merchant signs up to Cubyn and plugs in their various sales channels, such as Amazon, Rakuten, eBay, Shopify etc. They then send Cubyn an appropriate amount of inventory to fulfil future orders, which is stored temporarily in a Cubyn warehouse. When an order is placed, Cubyn automatically packs the order and ships via the most suitable carrier to optimise for transit time and cost.

“Our customers pay based on the number of parcels they ship,” explains Baca. “Logistics is a game of volume and thanks to technology we can manage volumes that couldn’t be managed by historical players. This allow us to offer… cheaper prices and still have great margins”.

Hot startup Superhuman has been getting some ‘backlash’ as happens now and then when someone notices the precise methodology that a startup is using to enable a really freakin’ cool feature set. We’re well into stage 2 now when, inevitably, the backlash itself gets backlash.

The nut of it is that people have been exposed to the idea that Superhuman tracks email you send and receive and allows you to keep better track of it. They do it on your behalf, but without the permission of the recipient.

You can read a review of the service by Lucas Matney, who spent six months with it, here on TC.

The best thing about all of this defense chatter coming in is that the backlash itself is really not all that serious. People are literally just pointing out what they do, which is track email. And it provides real, genuine value.

This isn’t, obviously a new idea. It’s done by every marketing platform worth a darn that uses email. Every single email that comes in from a BRAND has some sort of this stuff happening. As do all websites (including this one). People are just not used to it being applied to a consumer product as intimate as personal email, and that sort of in-your-face use of commerce-grade tracking is perking up ears.

A few years back a startup founder with a suite of productivity apps (not Superhuman) asked me about this cool new feature they were planning on shipping: email tracking for senders, built right in. Read receipts and action items and all kinds of cool sounding stuff to make your life easier. He was asking what I thought of it, and whether Apple would have an issue with it if they shipped it on the store.

I told him it sounded like a great idea, but that I would be very cautions of actually rolling it out because it was impossible to get verification from the other side before you began tracking them. There was no opt-in.

I advised him to look at the way Apple handles it, where email tracking happens outside of the body of the email in a sort of passive radar fashion. Instead of active ‘pings’ using tracking pixels or other image hosting tricks, you’re getting a lighter client-side data set to work from. It’s opt in on your side, and doesn’t extend to them.

I warned on it for the same reason that I opt out of services that route my work email through their own servers, I choose not to employ any tracking apps and set up my emails not to auto display images. It’s not because I don’t want actionable insights, its because I am unable to obtain the permission of the people I send it to to begin tracking them.

Yeah, for sure, they’re already tracked 10 ways to Sunday by every spam email from Groupon to The Gap, but this is coming from me, an individual. It’s different, in my opinion, which is why people are reacting the way they are.

Flash forward and now we’ve got a very well capitalized startup with this at the core of their business. It seems like the founders have thought a lot about this and have decided that this tracking is good and defensible. So it shouldn’t be a shock when it comes time to defend those choices.

If you’re a founder, I think that’s a core lesson: always be willing to die on whatever hill you’re building.

I don’t think that the chatter about the tracking feature of Superhuman is a case of people turning on a startup that has become successful. Superhuman is very new, but very buzzy. And, as I said above, the backlash mostly consists of people highlighting their marquee features in detail. I’d bet a lot of people became even more interested in what it’s doing reading the various and sundry tweets and posts about it, including a Big Profile post in the NYT that kicked off this latest round of discussion.

We’ve been covering Superhuman for a few years now, including detailed explanations of what they want to accomplish and what the origins of the product and team are. That’s pretty much our job — to make sure we see this stuff years before anyone else. Heck we even covered the last startup to use the name Superhuman for a productivity app. The tracking stuff has come up in our stories, but I think that people are just more willing to be skeptical of this stuff given the way that the last couple of years have gone. This is something that we have found happening with a lot of privacy issues recently.

In fact, the most astute criticism of the way Superhuman uses tracking came in a post by designer Mike Davidson, who has spent a lot of time working on large systems that have dangerous, as well as exciting, potential. And that post is anything but a ‘drive by’ on the model. It’s a thoughtful critique that actually offers some possible solutions.

I do think they are trying to solve a real problem. But there are clearly components of the way that they implemented their key feature that have potential for abuse.

It is, and I do find it a bit amusing that I have to say this in twenty nineteen, OK for people to want to discuss this and to examine the trade offs in a product that makes other people’s privacy choices for them. This isn’t backlash, this is discussion, and it’s good.

One of the reasons that we’ve gotten to a place where large platforms have been able to be mis-used to manipulate audiences at scale is that not enough people were listening to the conversations that were had about these possibilities early enough.

In context, it is very hard to argue that a genuine moment of thoughtfulness about any startup that has traction, raises significant capital and is aiming to have the most users possible see the world from its point of view is a bad thing.

Podimo, a Copenhagen-based startup building what it hopes will become Europe’s “Netflix for podcasts,” has raised €6 million in seed funding prior to launch. The round is co-led by Germany’s E.ventures and Denmark’s Heartcore, reflecting the young company’s two planned country launches later this year.

Founded by Morten Strunge, who has a track record in subscription media products via audio books service Mofibo (which he sold to Storytel), Podimo is hoping to capitalise on the rise in consumption in podcasts. Ambitiously, this will include both a free and paid version of its product, with the aim of creating a reliable revenue stream for podcast producers. The startup’s other founders are Nikolaj Koppel, Andreas Sachse and Sverre Dueholm .

“Podcasts have finally come of age and we are seeing a lot of demand for audio content globally across many different demographics,” Strunge tells me. “Consumers are increasingly looking for premium, ad-free services and we see a huge potential in the podcasting space”.

The Podimo app has been designed to provide a “superior experience” in discovery and recommendation compared to existing podcast streaming and download services. The idea, says Strunge, is to make it as seamless and easy as possible to find your next podcast.

“We believe that with the fast increasing amount of podcasts available, curation and discovery becomes more and more important to both unfold content in a relevant context and to the right individual user, which will benefit both podcast creators and consumers,” he says.

By launching a freemium model, where a paid version provides unlimited listening and features, Strunge believes there is an opportunity to work closely with podcast creators to strengthen the podcast ecosystem and make it less reliant on advertising revenue. “We want to become the preferred partner for creators, by both working closely with their content, curate and match it with each individual user, but also by offering a superior monetisation model,” he explains.

The hope then is that a more robust revenue stream will enable new podcasters to enter the market and existing ones to earn more. In turn that could give podcasters the financial headroom to invest even more time and effort into “creating great content”.

“Our dream is that with around 20% of people in Europe listening to podcasts on a weekly basis, many creators should be able to make a living out of creating podcasts, it shouldn’t just be for the few,” says Strunge, perhaps ignoring the fact that media often scales to become a hits-driven business. “We will offer revenue share to all existing podcasters out there, but also co-produce and produce original content,” he adds.

More broadly, Strunge says he remains a strong believer in audio as a format. He says not only is it easier to listen than it is to read but that podcasts are built for subscriptions. “It’s a short format, actuality driven, series driven and niche and broad at the same time,” says the Podimo CEO.

In addition, production cost are low so it is possible to keep to a price point below music and VOD services and Strunge is convinced we’ll continue to see a significant increase in the number of podcasts produced. This will include the broader market but also podcasts from more professional media players yet to invest strategically in the audio format.

President Donald Trump recently promised to ease the ban on American companies doing business with Huawei, but the Commerce Department is requiring its staff to treat Huawei as if the blacklist is still in place, reports Reuters.

Enforcement staff were sent an internal letter this week by John Sonderman, the Deputy Director of the Office of Export Enforcement, to continue treating Huawei as blacklisted. The letter, viewed by Reuters, said applications from companies that want to sell to Huawei should be considered on merit and flagged with language that notes Huawei is on the entity list. The applications should also still be viewed under a “presumption of denial” policy that applies to companies on the blacklist. This means license applications are scrutinized more closely and most of them are rejected.

Along with 70 other companies, Huawei was added in May to an “entity list” of companies that U.S. companies are forbidden to do business with. As a result, many of Huawei’s most important component suppliers, including Qualcomm and Intel, severed ties with Huawei, while Google cut off its access to Android--a major headache for Huawei, which is the third-largest smartphone maker in the world. Huawei founder and CEO Ren Zhengfei said the ban would result in $30 billion in lost revenue.

According to Reuters, this is the only guidance enforcement officials have received since Trump’s surprise announcement, made after he met with Chinese premier Xi Jinping at the G20 summit. In an apparent concession to China, which sees Huawei as major sticking point in the U.S.-China trade war, Trump suggested that the U.S. will allow American companies to resume selling hardware to Huawei as long as it doesn’t pose a “great national emergency problem,” and would hold meetings about Huawei’s trade status.

After Trump’s announcement, Ren downplayed the effect of the promised partial reprieve, telling the Financial Times that the ban has helped the company “become more united than ever.” He added “if we aren’t allowed to use U.S. components, we are very confident in our ability to use components made in China and other countries.”

2018 was the year when Didi Chuxing, the ride-hailing company that defeated Uber in China, vowed to put safety ahead of rocket-ship growth after two deadly passenger incidents. One way it works to ramp up safety is a stricter vetting process for drivers.

The mobility giant said Tuesday to a group of media that it has removed more than 300,000 drivers who don’t meet its standards since launching a safety overhaul last year. That adds pressure to an already loss-making company, which some speculate could be hit with a halved valuation (link in Chinese) since reaching an $80 billion high last year, for which passenger wait time is crucial to the success of its business.

In the long term, Didi could address the driver shortage by betting on robotaxis. The Information reported this week that the seven-year-old company is in talks with its largest shareholder SoftBank and other investors about raising money for an autonomous driving unit.

Didi is also hoping to keep both drivers and passengers happy by hiring more support staff. The company said it now has some 9,000 customer service reps on standby 24/7 to take questions from drivers and passengers; half of the agents are in-house staff. The number exceeded an earlier goal of 8,000 disclosed last September when Didi announced to spend some $20 million on customer service.

The hiring spree marks a moment of reckoning at Didi who had been fixated on collecting passengers and drivers while falling short in fulfilling social responsibilities attached to a tech platform. Case in point, Didi’s outsourced passenger support system was criticized for failing to act promptly on irregularities flagged by customers. Labor costs can swell, but safety measures have become a necessary business expenditure as the Chinese government call for more “social responsibilities” among private firms.

The same goes for other sorts of internet platforms like ByteDance, which runs TikTok and the popular news app Jinri Toutiao. The startup often touts itself as a “platform” distributing content to readers instead of a “publisher”. But following a series of online crackdowns ByteDance and other similar media businesses employed thousands of moderators to make sure their content is in line with laws and Beijing’s directives.

Didi claimed that it receives 300,000 calls from passengers and drivers daily, while only 1.7% of the inquiries are suspected to concern safety. It said its customer service system — which utilizes bots to solve standard questions — can react to 95% of the safety risks flagged under 30 minutes. For some context, China’s transportation regulator require safety issues to be responded within 24 hours and processed within five days, according to a set of guidelines published in 2016 to regulate the nascent industry.

TikTok is being investigated in the U.K. for how it handles the safety and personal data of underage users. According to the Guardian, information commissioner Elizabeth Denham told a parliamentary committee that the probe started in February, after the U.S. Federal Trade Commission levied a $5.7 million fine against TikTok for breaking children’s privacy law.

Denham told the Guardian that the commission is examining how TikTok collects private data and concerns about the open messaging system, which may allow adult users to contact children. “We are looking at the transparency tools for children. We’re looking at the messaging system, which is completely open, we’re looking at the kind of videos that are collected and shared by children online. We do have an active investigation into TikTok right now, so watch this space,” she said.

The investigation will also examine if the popular app, owned by ByteDance, violates the General Data Protection Regulation (GDPR), which requires companies to put special protections in place for underage users and provide them with different services than adults.

The FTC’s investigation, which began when TikTok was still known as Musical.ly, ruled that the app broke the Children’s Online Privacy Protection Act by failing to seek parental consent before collecting names, email addresses and other personal information from users under 13. The ruling resulted in an age gate being added to an app that prevents users under 13 from filming and posting videos on it.

ByteDance, the Chinese media startup now valued at $75 billion, told the Guardian in a statement that “We cooperate with organizations such as the ICO to provide relevant information about our product to support their work. Ensuring data protection principles are upheld as a top priority for TikTok.”

The new era of tech-enabled banks is coming, even in regulation-heavy Japan. Kyash, a fintech company with visions on becoming Japan’s first challenger bank, said today it has raised $14 million to continue its expansion.

To be clear, Kyash isn’t a bank. Yet. But it is currently applying for a host of licenses in Japan that could allow it to offer banking-style features including checking accounts, ATM withdrawals and money remittance. Right now, it is a payment app that offers a connected Visa card in the style of Monzo, N26, Revolut (which has a Japan license) and others of that ilk.

The startup was founded in 2015 in Shinichi Takatori, a former banker and management consultant who saw the potential to merge tech and finance.

“I really noticed that information and communication has become ubiquitous but money itself hasn’t changed for a long time,” Takatori told TechCrunch in an interview.

The company took some time — two years — before it released a consumer product, but it quickly tied up with Visa to offer a prepaid debit card that connects to the Kyash app. That provides benefits like instant payment notifications, clear balance and lower fees for overseas spending, while costs are born by merchants rather than users. They might seem elementary today, but they are still not standard among Japan’s traditional banks, Takatori explained.

The company declined to share its user numbers, but Takatori said that this new round of funding — Kyash’s Series B — is a validation of the progress it has made.

The $14 million investment is co-led by Goodwater Capital, a U.S. investor that has backed fintech startups like Monzo, Stash and Toss in Korea, and Mitsubishi UFJ Capital, the investment arm of Japan’s largest bank.

Mitsubishi’s involvement means that Kyash counts Japan’s three largest banks as investors, with SMBC, Mizuho having previous put money into the company. Others that took part in this Series B include Toppan Printing, JAFCO and Shinsei Corporate Investment Limited.

So many banks on the cap table might seem like a strange thing for a disruptor — let alone the banks, which tend to behave territorially — but Takatori believes that there’s the potential for cooperation, not to mention that it will help the startup with its licensing efforts. Already, he revealed, Mitsubishi plans to integrate its card with the Kyash app to provide its customers with the best of both worlds.

“We’re not here to win over existing banks, but instead inform [them of] how money should work in next decade,” explained Takatori. “So why not collaborate in some way.”

Kyash has a tie-up with Visa that allows it to offer its customers a connected debit card and also provide issuing services to other fintech startups

There’s also the fact that, even with a license, Kyash and others are unlikely to be able to offer full banking services. That means they will have to serve as complementary offerings to the industry, which would likely mean that cooperation is good — essential — for both sides.

But, beyond the consumer play, a notable piece of Kyash’s business that has investors excited is its B2B payment business.

The company developed its own payment processing system to reduce costs, which is one reason why it took time to launch. Thanks to a tie-up with Visa, it offers both issuing and processing of prepaid Visa cards to fintech companies in Japan that want to go down the payment route.

That’s increasingly popular given the government push to make the country a “cashless society” ahead of the 2020 Olympic Games next year. It could also appeal to crypto companies in Japan, which offers the world’s most robust licensing, who want to follow the example of the Coinbase card in Europe or startups like Crypto.com and TenX which offer similar prepaid cards.

Takatori said Kyash is “in discussions” with crypto companies, but that it has not made a decision on how to proceed yet. The company is also eying potential overseas expansions, although that is some way down the line.

“We have open eyes for globalization, it’s just a matter of when,” he told TechCrunch. “We still have a far way to go [in Japan, but] maybe after the Olympics.”

More pressingly, he sees the company looking to raise a “pretty quick” Series C round to give it acceleration into next year. That’s likely to go to more expansion and user acquisition since the licenses the startup has applied for are unlikely to be granted this year.

At Baidu’s Create conference for AI developers in Beijing today, the company and Intel announced a new partnership to work together on Intel’s new Nervana Neural Network Processor for training. As its name very clearly states, this forthcoming chip (NNP-T for short) is a processor built specifically for the task of training neural networks for the purposes of performing deep learning at scale.

Baidu and Intel’s collaboration on the NNP-T involves working together on both the hardware and software side of this custom accelerator to ensure that its optimized for use with Baidu’s PaddlePaddle deep learning framework, which will complement existing work that Intel has already done to ensure that PaddlePaddle is set up to perform best on its existing Intel Xeon Scalable processors. The NNP-T optimization will specifically focus on applications of PaddlePaddle that focus on distributed training of neural networks, to complete other types of AI applications.

Intel’s Nervana Neural Network Processor lineup, named after ‘Nervana,’ the company it acquired in 2016, is developed by the Intel AI group led by former Nervana CEO Naveen Rao. The NNP-T is tailor-made for training AI (ingesting data sets and learning how to do the job its supposed to do), while the NNP-I (announced at CES this year) is designed specifically for inference (taking the results of the learning process and putting into actions, or actually doing the job it’s supposed to do).

The NNP made its debut in 2017, and the first-generation chip is currently being used as a software development prototype and demo hardware for partners, while the new so-called ‘Spring Crest’ generation are targeting production availability this year.