Anodot’s automated business dashboards put big data in the hands of business people

bi and analytics, big data at Anodot

In many organizations, big data and analytics reside somewhere between the IT organization and marketing. That’s primarily because of the technical barriers around the stop-and-start nature of most analytics programs. To get an answer to a business question, someone — typically with some technical chops — has to write a query on the data to get at the answers.

That’s changing. For organizations to leverage the power of business intelligence and analytics, tools and systems are becoming easier to use and more real-time. Rita Sallam, research vp at Gartner, explained, “The BI&A market is in the final stages of a multiyear shift from IT-led, system-of-record reporting to pervasive, business-led, self-service analytics.” Once BI tools don’t require programming skills or a degree in statistics, their mainstreaming is practically guaranteed because they can add dollars to the bottom line.

One startup that’s taking this new approach is Anodot. I first heard of the firm in September (when it announced its most recent funding round), and I assumed the company was in cybersecurity because it does something it calls anomaly detection. But after speaking to David Drai, cofounder and CEO of the Israeli startup, it became much clearer that the company’s technology addresses big pain points in fintech.

Anodot’s machine learning looks at different buckets of data in an organization to establish a view of what normal looks like. So, if you’re processing credit cards, for example, Anodot examines approval rates across geographies to develop a model for what defines business as usual. Any deviation from the norm triggers alerts to relevant departments. The technology correlates abnormalities in the data to external events, like a new product launch, to see what’s causing them.

There’s no need to write hundreds of queries. “Our machine learning does this all for you,” said Drai, who previously cofounded Cotendo, which was bought by Akamai in 2011 for $268 million. “Anodot can answer two important questions for you: what’s happening and why.”

Credit Karma came to Anodot in 2016 with a specific problem. The company relies upon numerous funnels to ensure prospects and existing clients most efficiently interact with the firm’s credit scoring and monitoring service. All of a sudden, the revenue driven by a certain webpage dropped by over 50 percent over a three day period before the company could determine what was causing the problem.

“To test Anodot, we streamed a subset of six months of historical data to see if Anodot would find the same anomalies we had found manually, and it did,” said Pedro Silva, a senior product manager at Credit Karma. “It was clear very quickly that Anodot provides a ton of value to both business and technical teams.” After working with Anodot, Credit Karma quit further development of its own in-house solution to eliminate business incident detection latency.

Riskified is another fintech firm that uses Anodot’s real-time business dashboards. The company works with large retailers to prevent ecommerce fraud with instant approvals and full chargeback protection. Because it assumes the approval risk from its merchants, a break down in its risk models can be disastrous. The company looked to outside vendors for assistance.

“We think that someone focused on something specific will almost always do it better than generalists,” said Assaf Feldman, Riskified’s CTO. “Our customers know ecommerce really well but they turn to us for state of the art fraud protection. We’re all willing to pay for specialization.”

Tel Aviv-based Riskified uses Anodot to monitor internal signals of its fraud models and to make sure service levels are maintained for the product. The system’s alerts notify account managers if there are problems.

As Riskified looks to broaden its usage of Anodot across the company, fintech firms are taking a hard look at using new business intelligence and analytics tools in 2017.

 

U.S. Bank focuses on customer experience to generate direct marketing opportunities

Vanity metrics are dead. It’s social media data that drives insights in the age of big data.

U.S. Bank has a team that’s tasked with helping the company make data driven marketing decisions. The company employs an extensive social listening program that tracks US Bank’s Voice of Customer as well as the feedback on its closest competitors. With that input in hand, the bank makes investments in innovation, product improvement, customer service, and content and marketing opportunities.

U.S. Bank and big data

“Financial organizations need to move away from being gut driven and just doing things based on what others are doing,” explained U.S. Bank’s Troy Janisch, director of social intelligence, at Tradestreaming’s 2016 Money Conference. “Once you start listening, you can fine tune it to be more responsive to the products and services you have.”

Compiling the data is the first step in creating actionable insights for U.S. Bank. It averages about 28,000 mentions a month and has a single employee whose job it is to validate all these mentions manually. But events like the Wells Fargo fiasco can cause mention volume to significantly change month to month. That means negative sentiment about U.S. Bank can rise from a baseline of around 3 percent to upwards of 8 percent. To flatten out the variability, Janisch’s team prefers to use ratios, like net promotor scores.

The social intelligence team next weights mentions according to what marketing was working on during the month. So, for an awareness campaign, metrics like reach, share of voice, and visits will trump other engagement and acquisition metrics. By weighting what matters most, U.S. Bank can customize marketing campaigns to achieve specific goals.

One place this works really well is on sponsorships. The bank spends more money on sponsorships than it does straight-up advertising, so tracking performance is really important. Minnesota’s U.S. Bank Field is the National Football League’s newest stadium. To measure the impact of the stadium sponsorship, the social intelligence team benchmarks the Vikings’ stadium against other leading fields like Citi Field and AT&T Field.

“For each field, we can measure the social potential, how much activity is going on, and the awareness and engagement on each property,” said Janisch. “So we can set goals for U.S. Bank Stadium to be more effective than AT&T Field is for its sponsor.”

U.S. Bank also tracks local mentions of its network of 3000 branches. The social intelligence team knows that 70 percent of its customers and prospects will check online before making a financial purchase. To be more effective at a local level, the company tracks Trulia, Facebook, Twitter, Yelp and Google for mentions of its branches.

Once the firm has this data in hand, it acts by allocating marketing dollars where they’re most effective. For U.S. Bank, this is an ongoing process. The bank’s social intelligence team has made the firm more responsive by creating a continuous feedback loop to the rest of the firm on what’s working now, without getting bogged down in historical data.

“Dwelling on what happened in the past just slows things down,” explained Janisch. “Mass marketing is now more about continuous in-process marketing, creating systems as opposed to campaigns. Good creative is essential to compete, but that comes more at the end of really understanding the customer and being able to provide the right creative to the right customer at the right time.”

 

Hi 5! The top five fintech stories we’re following today

top 5 weekly fintech stories

Big data, big organizational challenges

Forget all the discussion about new user acquisition. How about focusing on servicing existing clients? Here’s how Vanguard uses data to deepen relationships with its customers. It’s not easy and requires a lot of organizational discipline, but there’s tangible payoff at the end.

Only 3 percent of banks claim they’ve created continuity across all customer touchpoints. It’s no wonder why digital efforts end up so fragmented. The old silos and departmental structures frequently impede change.

Top fintech podcasts

After we published our recommendations of the best fintech podcasts, readers joked that this wasn’t a best of list – it’s actually all of ‘em. Of course, implied in this bit of listicle-making is a request for you to check out our podcast.

(I)nsuring people stay healthy with Fitbit

Insurers are just beginning use consumer technology to encourage compliant behavior. For example, John Hancock’s Vitality program provides a feedback loop that encourages exercise. Using a Fitbit and a smartphone app, the insurer incentivizes policyholders to get off the couch and get moving.

I like big chatbots and I cannot lie

Ron Shevlin, Director of Research at Cornerstone Advisors, has been covering the financial services space for 25 years. At the Tradestreaming Money Conference last month, Tradestreaming editor Zack Miller had the opportunity to pepper him in a game of free association. Unscripted and unrehearsed, he riffed on things like chatbots, the future of the bank branch, credit unions.

Cashless societies

Physical money has been getting a lot of press recently, thanks to India’s recent move towards demonetization. There are lots of different views on whether the world will really phase out cash, but regardless of which side of the aisle you’re on, here are 3 stranger-than-fiction scenarios that wouldn’t pose a problem in a world without physical currency.

Selling Data: The emerging role of finance’s Chief Revenue Officer

new role of chief revenue officer

The financial industry is no stranger to data — both buy-side and sell-side institutions have a long history of buying data, even if it was only through a monthly subscription to a Bloomberg terminal.

As new business models appear in finance, revenue models are following suit. Selling data is becoming cool again and that means, for some firms, generating revenues becomes a priority.

We recently sat down with Jeremy Baksht, the chief revenue officer for Estimize. His firm crowdsources sentiment data around stocks (like earnings and revenue expectations) and sells this information to financial institutions. His title says a lot about his role and mandate — it’s not just sales. He’s also tasked with customer development, scaling the sales organization, driving adoption of paying customers, and, to a certain extent, contributing to product development.

Baksht, who joined the firm mid-2015, draws parallels between building a sales operation for a young financial data firm to his experience as an investment banker at Citi working with firms like GE, Northrop Grumman, and Carlyle in Asia and Europe.

He shared 4 key things for executives to focus on when scaling revenues in today’s financial markets:

Identify target customer segments

For Baksht, his job was somewhat clearer because when he joined Estimize, the firm already had a head of sales who focused on selling into hedge funds. Bakst’s job was to sell to other segments. And to do this, he had to quickly identify viable customers.

He soon targeted options trading brokerages as potential clients: “Large brokers are very interesting for us,” Baksht commented. “Our data can help options traders going in and out of strategies around earnings announcements. In this way, we’re a more modern whisper number. ” Estimize now counts 2 of the 5 largest options brokerages as clients. These firms count 6 million accounts in total with 1 million of them active.

The next category the chief revenue officer identified as a potential market segment was the sell-side. For sell-side clients, it became clear that the firm’s data could be a differentiator for equity sales teams. What Estimize has done is work with sell-side firms to become contributors to Estimize’s data as well as potential clients. Once a bank’s analysts are included in Estimize feeds, equity sales teams can see where their analysts are in relation to the rest of the market and use this to market their firm’s ideas. For now, sales, research, and trading (SRT) clients generally don’t pay to access Estimize data but Baksht doesn’t see a reason they wouldn’t become paying clients of the firm in the future. For now, he’s content to get the sell-side contributing to Estimize data and beginning to consume it as well.

Create beachhead clients

Baksht eventually closed his first deal with BCA Research which became his firm’s first non-hedge fund client. BCA is a nearly 70-year old independent provider of global investment research that focuses on macroeconomic data. By tying up with Estimize, the research provider can offer its clients an indication of certain stocks tied to a macro strategy that were most likely to beat expectations. Estimize is the micro to BCA’s macro strategy, providing actionable recommendations to its clients.

Estimize’s CRO also sees some challenges selling in to large financial clients. A market that is increasingly gravitating to index products is also quick to sour on bigger-than-life stock pickers. He cites Bill Ackman, who’s made billions for himself and clients over the years, but is having a tough recent bout with the market. “There’s a growing skepticism among institutional money managers,” Baksht said.  “We’re seeing more and more fallen angels crack on their own mistakes, making it harder to sell data [intended to beat the market]. It’s hard to see a human beating smart algorithms over time.”

Professionalizing the product

Baksht also found a way to work with large buyside firms like Steven A Cohen’s Point72. Buyside firms are interested in Estimize as a stock screening input into their investment decisions, but they’re also interested in the workflows the product team has created. For example, when speaking to a director of research at a fund like Point72, Baksht makes sure to emphasize that his software can be used as a compliance tool, as well.

In this capacity, Estimize can be used by a client to analyze its team of stock pickers. “We log the analysts and see where they fall out vis-a-vis their peers,” Baksht commented. “We provide compliance tracking and ranking to track teams of analysts. It’s a kind of paper trading — to see whether an analyst is right or wrong.” Baksht said some clients see his firm’s software as a development tool: if stock pickers aren’t doing their job to the best of their capacities, a firm can step in and coach performance.

Create incentives to build-in product reliance

It’s here that Baksht’s background in M&A helps him play the long game. His experience as an investment banker working on M&A transactions with industrial firms instills him with a form of patience to work through the entire sales cycle. “I’ve worked on deals with 5 year cycles and know how to be patient,” he admitted. “I get past common objections — they don’t bother me. I can grind it out over years, applying rigor, patience and discipline from other industries. I don’t have the same biases [as other sales professionals in financial data] — I have a fresh view and fresh legs.”

He’s confident that getting a firm to use Estimize will lead to good things in the future. Firstly, any firm consuming Estimize data, in any form, becomes a potential contributor of data in the future. If his company intends to professionalize crowdsourced data, getting more and more professionals contributing their data improves the product.

Baksht also incentivizes usage. He’s been willing to give sell-side institutions access to his product in return for contributing their analysts’ coverage of stocks. More, Estimize encourages professional clients to de-anonymize their estimates when they submit them to the earnings database. Being able to use an analyst’s name and firm around an earnings or revenue estimate is valuable for the data and Baskht has exhibited a willingness to find the right subscription price point for clients in return for their contributions. In this way, sales can not only drive revenues back to a firm selling data — it can also improve the product along the way.