Driving $$$ sales – can Twitter find the in-moment buyers?

News this week that Twitter is going to utilise the new Stripe Relay service has got a lot of people excited. Twitter users love to chat about the products they are seeking information on, looking to compare and contrast or out and out declaring they want to buy. All this chatter (or what we call pre-purchase intent) has a major blocker in the way if a purchase is going to happen – how to make a payment! If like the majority of Twitter users you are on your phone the hassle of dropping into a web site to enter your credit card info and set up an account is too much. Stripe Relay lets you do it all inside of Twitter on your mobile.

According to Patrick Collison, co-founder of Stripe only about 15% of e-commerce sales happen on a mobile phone whereas 60% of buyers are viewing the item they want to buy on their mobile. Stripe makes it easy for users to make that purchase in the moment; removing the friction for the user and enabling a transaction to happen. Stripe has made it easy to move in the moment from looking at something and making a purchase; hence why it has seen its valuation fly beyond $5 billion!!

All the excitement about Stripe and Twitter teaming up is well warranted; the opportunity for Twitter to improve its overall conversion rate is clear. Whereas Facebook remains the dominant social network for conversion stats Twitter does languish somewhat with 0.77%.

Twitter’s Head of Commerce, Nathan Hubbard certainly is excited at the opportunity. According to him Twitter has “50 million-plus users every month who are tweeting some signal of intent with the words ‘I want’ or ‘I need’. By providing these users with a more frictionless payment experience Twitter sees a game changing uplift in social commerce revenues.

Why be so vague? Twitter is retrospectively looking at user purchase intent and ‘hoping’ that that these users will now transact. The words ‘I need’ may relate to a purchase or could be ‘I need to leave Twitter’!! You can’t accurately view signals of intent across such a vast amount of Twitter messages without a scalable solution like ‘Path 2 Purchase’ – our machine learning solution created over 6 years in academia.

The real game changer is to accurately understand and find purchase intent in the moment – as in REAL TIME.

Google’s research on the proliferation of ‘moments’ in consumer path to purchase is highly accurate; lots of ‘moments’ are happening as consumers dive in and out of the journey. One ‘moment’ to another ‘moment’ could be looking for a very different item or brand; the intent needs to be understood right in that exact ‘moment’ and action needs to be taken. Only machine learning can at scale accurately understand the right context of each and every moment and process a message for action.

We baselined our Path 2 Purchase machine learning solution on the Twitter firehose; we classify 148k messages a second. We don’t just find intent; we classify 3 distinct stages of intent (interest, consideration and confirmed sales lead) and can segment by industry or brand. The game changer is to match intent to action by brand and to drive the consumer path to purchase through to successful sale.

What would that be worth to Twitter’s share price??

If Twitter is now going to drive conversions it needs to find those 50+ million users a month in real time and successfully increase the conversion rate. Finding in-moment buyers is a different proposition to identifying consumers considering a purchase via keyword search and an overnight batch process.

For more information on Path 2 Purchase or our NLP Framework for creating disruptive and highly scalable machine learning products please visit our website www.chatterbox.co or contact Andrew Watson VP Strategic Alliances Andrew@chatterbox.co