ubisend was spun out from a larger company in 2016. At the time, mobile messaging was creating a huge impact in the lives of consumers. We began spending more time using messaging applications than social media applications like Facebook or Twitter.

The ubisend founders saw an opportunity to create a new and exciting brand to meet this new demand. With most mobile messaging applications opening their APIs to third parties, ubisend could now build their super smart solutions straight into the digital channels like Facebook Messenger, Telegram, and more.

ubisend provides truly multi-channel service to its extensive list of Fortune 500 customers

ubisend stands for ubiquitous sending. Our platform is unique in that it centralises all messaging applications into one. We get to build a so-called ‘chatbot’ (we hate that term) for one messaging application, say Facebook Messenger. Then, at the flick of a switch, make the same chatbot available simultaneously across 28 other channels like Telegram, email, SMS, Line, etc.

We also are in a unique position, working with some of the biggest brands in the world. As leaders in the industry, we pioneer many conversational techniques, technology applications and user stories. As such, we don’t provide a one-size-fits-all product help; we offer bespoke, high-end solutions to unique company problems.

Current state of bots

Current limits of bots so you know when to use a bot and when not to

To me, the only real limit to bots at the moment is empathy, a bot cannot truly be empathic. This means even if the technology allows it, I would advise a business not to use a bot for certain parts of their operations. For example, responding to distress calls or severe mental health issues could probably be handled by a bot, but I would not recommend it.

How to avoid screenshot worthy mistakes by your bots?

Working with such large brands, we have to be super-careful about this sort of thing.

The first thing to realise is most large enterprises begin adopting this technology internally first. This is often the reason why ubisend is not allowed to share information about the work we do. These internal, employee-only, bots are not accessible to the public, which obviously reduces the chance of ‘screenshot worthy’ events.

Then, technically speaking, we are constantly running tests and optimising the natural language understanding. From the very first part of the build all the way to delivery and through maintenance, we’re always testing the bot. This allows us to catch most issues before anyone even sees them.

Finally, big companies are just not willing to let a machine (regardless of how well it’s ‘trained’) talk their customers. Their brand, values and bottom-line are too important and closely guarded. Because of this, most large businesses deploy solutions using a custom language matrix. We use machine learning techniques to automatically create a language corpus (and personality) which a company’s PR, legal, marketing, and whoever else, can approve. These types of bots will still keep learning over time, get smarter and smarter, but adhere to the approved data and language. It won’t ever be able to swear or be rude to users, simply because these words are not part of the approved language matrix.


Pricing a chatbot is a little bit like pricing a website. If you want a simple do-it-yourself WordPress site, you can probably get one for $50. If you want a complex, e-commerce, multi-lingual, Salesforce-integrated, CRM-driven online platform, you will probably need something bespoke and find experts with a proven track record.

For chatbots, we are those experts.

A custom build and implementation generally requires an investment of around £10,000 and can go up to how-long-is-a-piece-of-string, depending on what the customer needs.

Best practices for Fortune 500 that want to start with chatbots

I would give the advice I always give: start small, with a focus.

Most of the time, companies come to us with an idea or problem. During our first chat, when we say we can basically build a machine-human that can do anything and everything they need, their eyes widen and all of a sudden, they want everything. I would say one of my biggest jobs in any meeting with a Fortune 500 is to rein everyone in and get them to focus on one task. Treat it like an MVP, start small, prove the business case and invest over time. No one in the world knows the best solution for their problem, it’s up to us to collaboratively figure it out with them. So, we focus on the task at hand.

What are we improving and why?

Start with the what and the why. Everything else is done step by step later, we first need to nail the one specific issue.

As far as the best ROI, it’s quite hard to pick just one as it’s dependent on the company’s profile. At ubisend, we have seen great results in the HR space. It’s a fantastic test bed for businesses new to artificial intelligence. It’s company-facing (less risk of that screenshot-worthy moment!), it’s central (everyone gets to interact with it, from new hires to CEO), and HR is often an afterthought in terms of company optimisation (therefore, needs the most help!).

Starting with HR makes a lot of sense. We quickly improve internal processes and communication from the inside out, making an almost overnight impact before moving on to other departments like sales or marketing.

Best practices for SMEs that want to start with chatbots

SMEs will soon have all the power. The democratisation of technology is a real, and very soon business of all size will have access to all the fancy things normally reserved to a big-budget enterprise.

My advice is to stay in the loop. By 2020, 80% of businesses will have some sort of chatbot implemented (http://uk.businessinsider.com/80-of-businesses-want-chatbots-by-2020-2016-12). Use these two years to educate, start prototyping, stay aware of the opportunities and perhaps test a simple off-the-shelf chatbot.

At ubisend, we do a lot of work with SMEs. We contribute to the democratisation of the chatbot technology. We truly believe conversational AI is the future, and this future will be powered by enterprise companies and SMEs alike.

We also work hard powering technology for non-profits. NGOs, charities and people that can prove they’re doing good in the world can use our services for free.

5 predictions on the future of chatbots

1- In the tech space, I believe we will continue a rapid growth in the development of tools that allow consumers to benefit from the power of bots and 24/7, machine-led communication. The tech will become more common, less expensive, and more intelligent quicker than we think. This will widen the opportunities for businesses of all sizes and budgets to get involved.

2- In the UX space, I believe any development will help. Interestingly, there isn’t much out there yet in terms of ‘chatbot UX specialists’ or ‘conversational specialists’. We will see huge changes in this field. UX designers and creative agencies are gradually creating the niche and can’t wait to get involved. Conversational UX is far beyond words, colours and layout. It’s now all about conversations, tone of voice, and language.

3- The future of chatbots is bright. Soon, we will move away from the word ‘chatbot’, because it’s gimmicky and doesn’t represent what this technology can do. Hopefully, we will settle for something a little better, like conversational agents or conversational software. I think voice will take a big slice of the action in this space, but, will remain focused on more private, intimate conversations. We already have access to voice (Siri, Alexa, etc.). Yet, many of us still find it awkward to engage with these agents through voice, especially out in public.

4- I’m most looking forward to interconnectivity between bots. Once chatbots become mainstream, the next step will be ‘chatbot API’ing’, for a lack of better term. This API layer will connect chatbots to other chatbots and allow seamless experiences and data transfer throughout our day.

5- Finally, in a not-so-distant future, I see two major tech trends merging: blockchain and conversational software. At the moment, consumers are often concerned about sharing personal data with a chatbot. Perhaps, one day, we’ll be able to monetise and monitor our data through a decentralized blockchain. No one will single entity will have the world’s personal data, we’ll all own and be rewarded for the data we decide to share.


I want to thank you and AIMultiple for the opportunity to share my thoughts. I know the chatbot (conversational software!) and AI space can seem scary to some. If you are interested in this technology, I urge you to come and read our blog. We make all the information easily understandable to anyone, regardless of your technical geek-ery.

Come hang out with us!

Contributed by Alex Debecker, founder and CGO of ubisend.com, a creative AI technology company based in the UK. ubisend is one of the leading conversational AI companies in the world, working with companies like Unilever, Johnson and Johnson, Babycentre, and more

How useful was this post?

Click on a star to rate it!

As you found this post useful...

Follow us on social media!

How can we do better?

Your feedback is valuable. We will do our best to improve our work based on it.

Leave a Reply

Your email address will not be published. Required fields are marked *