Personalisation at Scale: Twilio’s Signal London 2024

Twilio’s Signal event series brings the company’s customers together around the world. In November 2024, a large event full of exciting demos, customer talks, and announcements demonstrated Twilio’s commitment to its customers.

Hundreds of Twilio customers made their way to Central London, to a great venue on The Strand to hear the latest news from the exciting CX company. The main theme of the day which CEO, Khozema Shipchandler, focused on during his opening keynote was ‘Personalisation at scale’, and how Twilio was working to personalise every single customer journey that it manages. He also discussed the power of Rich Communications Services (RCS) and how it would help combat fraud by allowing carriers and brands to guarantee the authenticity of outgoing communications.

How do we ‘Personalise at Scale’?

The concept of ‘Personalisation at Scale’ is a strong theme for Twilio to focus on. One of its key differentiators as a company, is the strength of its Customer Data Platform, and properly maximising this in 2024 naturally leads to enhanced personalisation. This ‘strength of CDP’ was re-emphasised by the first customer that presented at the show, Italian Luxury Clothing Brand, Luisvisaroma who had transitioned to Twilio to build a CDP and a unified data infrastructure to service its large online retail business.

To emphasise this point, Twilio demonstrated lots of different scenarios where customer personalisation could happen, such as providing that data to agents, or enabling automated customer journeys with that data. An example used a few times during the show was of a frequent traveller that is staying at a hotel that they regularly visit. In the examples, the hotel could automatically provide coffees, breakfasts, book taxis and provide other addons that recognised the guests’ preferences and habits to provide a consistent personalised experience.

This message was re-emphasised by the CMO, Chris Koehler, who spoke about building a unified customer profile that included all data on a customer, as well as what channels they preferred to be communicated with. He stated that in modern CX, customers have very strong preferences over what channels they use, so that brands which can respond to this will drastically differentiate themselves.

There was a slightly awkward moment, where Chris (who was keen on audience engagement) asked the audience who had a data warehouse, and very few people raised their hands. This did raise the question of whether people in the audience were just hesitant to engage, or whether many of them have their data hosted in cloud datacentre providers and do not feel like they have a ‘data warehouse’ in the traditional sense anymore.

The next speaker Kathryn Murphy, SVP of Product, ran through a quick demo of Virtual Agents and Agent Co-Pilot, all of which are in beta now. These demos were fine and roughly in keeping with what other vendors have been working on, but as they are still in beta while other vendors are starting live deployments already, there is some indication that Twilio needs to catch up with some of its competition on the AI front.

Also interestingly on the AI front, is the choice to power their AI assistant with an OpenAI integration, when many other industry peers are choosing to deploy non-open model-based systems that draw on much more limited datasets rather than free-flowing generative AI based systems. However, it seems in practice customers are still focused on building fixed response agents, as was highlighted by EON Energy who spoke during the afternoon and talked about the risks of free-flowing conversational agents. However, as EON pointed out LLM integrations were useful for intent extraction and understanding customer communications even if the responses allowed were fixed.

This was further reinforced by the other live demos on the day. I will admit to being a little underwhelmed by the ‘Live demo’ of their conversational AI engine. For a live Demo on a keynote stage for their largest customer event it seemed a little under polished. Long response times, obvious errors like suggesting getting an underground service from a station that did not have the service, and a fake feeling attempt at empathetic communication clearly demonstrated how far this technology still must go. The much better demo was the one that was demonstrating specific interactions handling a bank fraud check, where it was clear that certain interactions had been preprogrammed. I’m certain we will end up with GenAI powered conversational AI shortly, but I don’t feel that we are closer to that goal yet.

The strength of Twilio’s CPaaS offering was still on full display at the event, especially in the talk from one of its customers Fresha. Fresha is a communications provider to the personal care industry including hairdressers, masseurs and other similar businesses. The Fresha platform allows people working in this industry to manage employee schedules and automate booking management.

It provides these customers with a clear platform for the full management of their business day-to-day, as well as directly handling many of the pain points that might emerge like no-shows by sending automatic reminder messages before the appointment.

For those not in the industry this might seem like a simple thing, but its important to not underestimate the impact a smooth system like this can have on an industry where even a franchise of a large organisation might still only have 4-5 employees working. The employee handling the management is not likely to be a fully time manager and reducing the amount of time they have to spend handling admin can greatly increase their revenue generation. I think this a great case study as it brings home the real world impact of the Twilio platform, and highlights the businesses strength at providing effect CPaaS to a large number of businesses across the globe.