This morning, as I was drinking my coffee, I happened upon this article. The philosopher in me couldn’t stop reading it.
The public debate over LaMDA is often pointless and highly sensational: Is artificial intelligence (AI) sentient? Will AI escape our ability to control it in the future? Not to mention a ton of other questions heavily based in science fiction.
What people need to understand is that AI is impressive and it would be silly not to take advantage of it. But as to all the rest, it’s just a lot of noise about word usage that varies wildly from one side of the argument to the other.
The technology is advancing at a rapid pace and will become widespread a lot faster than has been anticipated. You won’t want to miss this evolution, because the digital industry of the future will be dominated by intelligent virtual agents!
As far as the use of AI within websites is concerned, I have a love-hate relationship with most of the big ecosystems I visit. The reason GAFAM experienced such explosive growth is because they responded decisively to web users’ need to efficiently find what they were looking for.
In 1998, Google made its sensational entry onto the scene because suddenly they made it possible to find everything you were looking for, and more. This was followed by the widespread use of the word “to google.”
I’ve been giving talks on managing personal data and digital audiences for close to eight years with Infopresse. Do you know what the recurring theme has been all these years? How companies can learn to respond faster to web users’ needs, even to the point of predicting what they’ll need and providing it before they’ve even landed on your site!
Google Now was a monstrous success at this. From an experiential point of view, web users no longer search for information. They receive it, and the platform predicts which articles will be of interest. The information scrolls in front of them and, if it corresponds to a need, they stop. It’s the same thing for Facebook, Twitter, Instagram, TikTok, BeReal and all the rest.
I think that websites in the future will be smart conversational agents that will surprise you with their ability to emulate the empathy of a good business representative! And yes, I’m talking about nothing less than an industrial revolution for sales assistants and websites.
The technology is there. It’s not that complicated to set up and the economic benefits will be interesting. And just to let you know, Desjardins has already integrated conversational AI into its phone system to reduce the cost of its call centre.
“[This virtual entity] is able, with just a single sentence rather than four or five automated options, to look after thousands of customers every day. It can replace a lost credit card, request a personal loan or an RRSP, change an address or reset a password. And it understands Quebec French” (La Presse, 2022)
The feed, pitch and products that will be showcased on websites in the future will not be predetermined like the pages of a catalogue. They will be “blank canvas”-style sites, like your Google Now feed, or like your Facebook page, or Amazon.
You might be thinking that your website is not the same thing—I would agree. But that doesn’t mean you can’t take some inspiration from the idea. Let’s return to the basic question: What are you managing and offering to your customers online?
If you have an idea of which products to highlight or pitches to promote, an ability to track strong trends, and a good understanding of your customer types, you can change part of your site into an automated selling agent. And you can still be sure that you’ll be in compliance with Law 25, which is a whole other subject unto itself. In fact, your ability to improve your marketing experience during this transitional period will be an asset when it comes to standing apart from the competition.
In this scenario, which I’ve purposely simplified, the web user lands on an empty page, and you will then have the following two options:
1) You have some data on the visitor and their authorization to use them, data such as the source of the visit, the medium of arrival, the terms used, the theme that led them to visit your site, your internal CRM data and proactive market analyses;
or
2) You have no data on the visitor, but you still have your business knowledge.
If the web user refuses the storage and use of their data, they will basically find themselves in the second scenario. However, without getting too technical, this doesn’t mean that the visitor will be completely lacking in indicators while interacting with you during their visit. What matters is that, at the end of their visit, the web user maintains the same level of anonymity as when they arrived.
Taking that as a basis, it’s possible to offer the beginnings of a personalized experience. Start with an introductory paragraph about the company, an image, an infographic or even a video. Explore this headless approach to doing business, keeping in mind that this isn’t necessarily a completely chaotic formula.
You still prepare your content, have a good command of your basic pitch, your paragraphs, your texts, but what changes in this type of approach is that the content unrolls based on the user’s exploration instead of sending them pages within an entirely predefined browsing architecture containing pages and pages of content. For example, the user could scroll down to the bottom of the page to deepen their experience.
The more the user’s engagement grows, the more your understanding of their needs is refined and the more your ability to predict how to solve their needs materializes.
Yes, this requires a lot of computing power. It’s not like everyone will be ready to jump into this future scenario tomorrow morning.
But with the arrival of Google Analytics 4 (GA4), using BigQuery for all digital interactions as well as Google Cloud Platform (GCP) (and more specifically, AutoML) will be a solution that comes pretty naturally. And don’t forget that now almost all content management systems (CMS), including those available for free, offer headless management. So it won’t be long until sites change their approach in this sense.
So there you have it.
To sum up, websites in the future will become intelligent agents.
A website will be a dynamic experience that is practically unique. The agent will learn who is visiting based on traffic—from every session, abandonment and failure. In short, from every interaction! What’s important is to ensure it’s well trained with all these data.
Meanwhile, in a data centre somewhere in the world, the site you’re interacting with is actually being managed by a collection of brain cells in a damp, warm little dish located on a chip.
Bizarre, or promising? There’s no definite answer, just many shades of grey. For my part, I love this kind of hard sci-fi thinking first thing in the morning.
If this has inspired you to reach out for a more in-depth conversation about this kind of project, we’re open to discussion and to launching concrete initiatives. Tackling this type of change is easier than you might think.
Don’t hesitate to get in touch!
