Why the next BI tool will be a data marketplace

There is no BI tool on the market that makes data exploration effortless. Cluvio and Mode want your organisation to learn SQL. Tableau, Domo, Chartio, Amazon Quicksight and Microsoft PowerBI try to hide the SQL behind a UI. You can use natural language in ThoughtSpot, but you have to learn the exact phrases, and the column names. Excel is still the most intuitive solution, but gives up around 1 million rows.

The problem is the same: the tools have a “data first” design, but our minds are not structured like data.

The problem is the same: the tools have a “data first” design, but our minds are not structured like data. Even an expert data scientist needs time to get to know the data. Strategists do not use the tools because they are not technical (in my experience, they often are). They don’t have time to learn the data model. A retail executive needs to understand supply chain, international price fluctuations, omni-channel marketing. When do they have time for data science?

BI needs a game changer. Imagine that if every time you wanted to search the internet, you had to ask a data scientist to build a dashboard. Then imagine your favourite search engine came along.

Data marketplaces are going to be big

The market for big data is growing fast. Real estate needs big data to price property in real-time. Retailers want to identify fine-grained customer behaviour. Finance needs to discover new alpha in the market. Experienced data vendors (Experian, Axciom, CACI) are ready, but there are challengers. Business analytics software vendors (Esri, Qlik) are bundling data into subscriptions. And new startups are building domain-specific data marketplaces (SkyWatch in satellite imagery).

Data is a difficult commodity to sell, because the ROI takes time to materialise in any business. The current process is long and complex:

(i) Find the correct representatives on the buy and sell side.

(ii) Identify precise requirements for the business.

(iii) Send over sample data, to check the structure and format will work.

(iv) Start a trial period with real data to measure ROI and justify a long-term commitment.

(v) Negotiate a long-term contract. This is unique as each customer has a different use case.

And this assumes that the data vendor has enough credibility for the quality of the data.

A business can avoid this if they have enough money (Hedge funds, multi-national FMCG). They buy the raw data and absorb the costs of any failed projects. But this prolonged process will struggle to touch the rest of the market.

To solve this, data marketplaces will discover a way to close the cycle faster: add analytics. Imagine a world without car dealerships. You would search online for car vendors. Sales representatives would meet you and bring car models one by one. You would pay a trial fee to test it for a couple of weeks to check if it works for you. Then you would negotiate a price. This drawn out process gets cut down to even one afternoon in a dealership. You arrive, sample cars, take a few for test drives, and sign the paperwork at the end of the day. This is “show-rooming”. And for data, show-rooming means trying out the data in an analytics settings. The customer needs to see how the data performs on their day to day strategic business problems.

For data, show-rooming means trying out the data in an analytics settings.

How the data marketplace wins: bring the analytics to the data

Soon the customer will start asking: what does this show-rooming tool offer me, that my in-house BI doesn’t? The data marketplace cannot compete feature to feature with Qlik or SAP. But it doesn’t need to. The engineering team will augment every data-set with meta-data and semantic information. And the user experience team use this information to design an interface that is natural. It’s the difference between “age is a numeric value in this column” and “age is how many years a person has been alive”. Soon, using AI, the user will ask natural questions, and the system will translate that into the data model.

This will be a game-changer for how decision makers can access insights. It will trump every other feature. It’s the search engine experience: have a thought, and get an answer, on demand. No time wasted with “how do I translate my human way of thinking to the software way of thinking”.

Centennials are a generation who don’t remember life without internet.

Centennials are a generation who don’t remember life without internet. They have an attention span of 8 seconds. This workforce will use agile, search engines, and chat. And they will get insights from an intelligent business insights tool, built on the data.