Startup of the week: Tekstum

Startup of the week: Tekstum

Sure, your book might get lots of likes - but what do the people behind the likes actually feel? This Spanish startup's new tool promises to analyse and visualise readers' emotions in real time.


The pitch

Tekstum claims to be the first technology to provide real-time scientific analysis of the opinions and feelings that readers are expressing online. Using a state-of-the-art algorithm built with big data and artificial intelligence, the company promises to revolutionise how publishers market, and how customers search for and buy books, by synthesising, analysing and visualising readers' moment-to-moment emotions and experiences.

Who's behind it

A multidisciplinary team. Spanish founder and c.e.o. Marc Santandreu worked for international publishing companies such as Author Solutions and Regió7. Lauren Romeo (PhD in Computational Linguistics) is the Lead Scientist, pioneering the Natural Language Processing research and development of the engine, while Juanjo Fernandez is the engineer responsible for the development of the software.

What's the gap in the market?

Every year the number of new books in the market grows exponentially, generating huge amounts of data at ever-faster rates. However, much of the analysis currently being conducted to help publishers, retailers and readers navigate this flood of content is quantitative (such as the total number of “likes” of a book) rather than qualitative (such as why someone “likes” or “dislikes” that book). Step up Santandreu.

"With the number of books being published constantly increasing, it becomes more and more difficult for a reader to use traditional means to discover new material," he explains. "Tekstum breaks through this wall of content and its limitations by not being based on past sales, searches, or the inaccurate self-evaluation star system but on the scientific analysis of the opinions and feelings that readers are expressing. Combining both artificial intelligence and big book data, we provide recommendations that go beyond simple lists of best sellers, editorial-provided metadata, or stars that do not contemplate the very fine nuances that could exist between stars."

Success so far

The Tekstum team took their product to market in February 2016 with a focus on the Catalonia region of Spain and has since secured 2,645 users. "Catalan publishers of all sizes have energetically embraced the Tekstum technology and its possibilities to shape their marketing strategies and campaigns" Santandreu asserts. "The next quarter will see a marketing push for the rest of Spain and Latin America."

Biggest challenges

Growing fast enough, before competitors emerge to ape the tech. "Over the next few years, our goal is to become the reference for emotion and sentiment-based book recommendation systems, as well as to continue expanding our engine to encompass more and more languages," Santandreu says.

Ultimate ambition

The Tekstum team is currently completing the testing phase of the engine for the English language, which will be released this summer. "It will open doors to new and exciting markets for us," Santandreu enthuses. "After the release of the English version, we also plan to expand in order to cover more languages; French and German are the next two on deck."

Advice to other publishing entrepreneurs?

"Passion, perseverance and focus."