Agents of Change

What Happened So Far

AI taking off has always been a prerequisite for the grand vision of CM.com to become a reality. In our September 30th 2019 prospectus for our IPO, this somewhat boring, yet accurate description of that necessity can be read:

“AI is expected to have a considerable impact on the future of conversational commerce, as it can automate and optimize routine processes through generating and analyzingvast amounts of data. One specific application of AI to messaging channels, which is expected to be at the core of future B2C communication, is the so-called chatbot. Through established chatbots, vendors can provide consumers with a continuous support function without the need for human capital.”

Though Generative AI wasn’t really a thing back then, it goes to show we’ve always trusted that the whole AI thing would work out – and in fact was needed for us to get where we wanted. Why wouldn’t a computer be able to do all these things that now require knowledge work? The road wasn’t clear yet, but the destination definitely was. In June 2020, we followed through on that awareness with a sizable acquisition in the Conversational AI Space. Other acquisitions and organic R&D efforts followed in the Contact Center and Marketing AI domain to piece together our AI powered, end-to-end- Customer Engagement Platform.

By 2022, we had arguably attained one of the broadest product portfolios in the world. A position we still hold today and a reality that we were, more than once, criticized for. “Why not just do one thing really well? This A2P SMS thing seemed to be paying off.” “Competitors are more focused” and other variations of the same comments were commonplace as we spent time and effort to bring it all together.

We knew back then, as we know today, that the market would look for consolidation in customer engagement software – and more broadly – Conversational Commerce.

Our software was strong, and each stand-alone product gained notable analyst recognition and accreditation. The integrated combination, however, was where the magic was and still is today. On the AI front, not just at CM.com, but globally, things were still “clunky”. Helpful for sure, but also clearly still in its infancy. In December of 2022, the ChatGPT moment was a wake-up call for many people to realize what enthusiasts had already seen unfolding:

AI should really be able to do anything. And eventually will.

Through our earlier acquisitions and other moves, we were bolstered with Data Scientists and similar enthusiasts. We were quick to move on this watershed moment with the launch of our RAG-platform, fully integrated with all our SaaS products. We offered “Chat with your PDF” before ChatGPT itself did. We’ve continued pushing since, with our big February 6th product launch being a testimony to our capability to stay ahead.

In Q1 2023, we also shared a new 5-year Product Strategy with 3 phases internally, that has served us well since and will continue to keep doing so. In this write-up, we want to share that strategy with you too and explain some other things as well:

• How we finally see everything falling into place.

• What to expect from us as an AI-first company.

• And why to be along for the ride.

Putting together an annual report requires some planning, so what you’re reading actually predates our February 6th launch event. A lot will happen between now and then. The AI landscape as a whole and CM.com specifically will be all about “Agents and Agentic AI” in 2025. We prefer to define the complete picture as Agentic Systems.

This is the second phase of our 3-step plan and it involves us widening our aperture from AI Applications for Customer Engagement, to AI to operate your business as a whole. Two years ago, we already anticipated that the market would gravitate towards an AI-powered best of suite approach for their entire application landscape. By now, this is the undeniable reality in the market. Almost every customer, every prospect, big or small, across industries – is moving from complex multi-vendor landscapes to single vendor preferences. We’ll review the 3-step plan to infuse and entangle our broad portfolio with AI more visually by the end of the piece, but let’s get back to our excitement for Agentic Systems for now.

What’s Next

In the two years that followed the ChatGPT moment, the popular press oscillated between the two sides of AI clickbait. From “AGI is coming this year”, to “The party is over, it was all a hype”, every couple of months or so. Both got eyeballs – and both were false. As was to be expected based on Amara’s Law, technology is always overestimated in the short-term and under-estimated in the long term. In 2025, at the start of the Agentic Era, we’ll all come to realize however, that this short-term is over now. As the use of pretrained models in organizations crosses over from text-only, vanilla “Retrieve and Augment” a.k.a. human-like chatbots, to multi-modal Agentic Systems that can perform actions, the B2B world will have its aha-moment, that consumer has already gone through.

Perhaps it’s a little “Captain Obvious” to write down, but getting stuff done is how economic value is created. If we just sit around and chat all day, nothing happens… So, even though we’ve already succeeded at making companies far more efficient with our offering, it’s easy to see how AI systems that can actually DO things instead of just chat, have a role to play in true AI adoption within organizations.

Organizations that are faster to adopt technology with deep impact, will see efficiency and income gains that allow them to lower their prices for their audience, as well as increase quality across all disciplines. In a global and open market, that forces everyone’s hand to do the same, or fade out of existence.

This new era in which AI systems can act, feels like the final piece of the puzzle to achieve what CM.com has been aiming for – for many years since its founding. The goal has long been to create one platform that weaves together Communication, Commerce and Customer Engagement. As we lay this final piece, we're also excited to embark on the broadening of that puzzle towards internal automation as well. To achieve that what our CEO Jeroen would phrase as:

“CM.com, the last software you’ll ever buy”.

So a few trends, anticipated as part of our strategy, are coming together in this confluence for us a company. Let’s try to unpack them a bit more extensively, as they are key to our excitement for the coming period.

Consolidation of Software

In 2011 Marc Andreessen (Netscape, Andreessen Horowitz) published an essay called

“Software is Eating the World”

In the B2B world, this insightful read was followed by the SaaS boom. The switch from legacy tools running on prem to SaaS led to scattered, multi-vendor application landscapes, as best-of-breed was pretty much the only choice you had back then if you wanted quality. Fast forward 14 years – and that is no longer true. The market is consolidating rapidly and the application landscapes within companies are gravitating towards single vendor approaches.

Organizations are waking up to the fact that they have overblown procurement, IT, and Business Intelligence departments, as well as an oversized accounts payable department, that exist at the behest of the upkeep of their complex application landscapes. Buying tools, integrating them horizontally, reporting on everything that happens within these tools, paying the bills… It’s all busy work.

It also makes you slow and siloed.

Best-of-suite is becoming the name of the game. If you want to be nimble and control costs, this is where it’s at. Our products as standalone solutions have already won awards and recognitions from Juniper, Everest Group, Frost and Sullivan, Forrester, IDC and Gartner alike. These are not “pay-to-play” stickers. You get into reports like these by sharing and showing what you can do in extensive RFIs and demonstrations that go on for hours and oftentimes include reference calls to clients as well. Though we don’t lead the market on all fronts, we feel it’s safe to say we make top-of-the-bill software. It’s when we combine these capabilities that have always been judged separately, however, that's where our magic really begins.

We barely see RFPs for stand-alone solutions anymore and the number of products that we serve per account is on the rise. CM.com has been leveled with critiques in the past that it was spreading itself too thin, that its product portfolio was too broad… That we may be better off just doing one thing really well.

We’re hearing that a lot less these days.

“Right-sizing” organizations is making the enterprise descend to other Independent Software Vendors

A very small group of players in the market, primarily those that cater to large enterprises, have portfolios that are equally broad to ours. However, you can’t get anything done there without consultants. The user interfaces and APIs are clunky. The prices are high.

Believe it or not, this all used to make a lot of sense. If you’re a big company with 100.000 employees, and perhaps a similar number of processes, it’s far cheaper for you to spend loads of money on highly customizable software, implemented by consultants, that fully adapt to your way of working on all fronts. The alternative of having all your people and processes changed to fit newly purchased “standard” software has always been more expensive if you looked at total cost of ownership.

However, with the advent of AI-driven efficiency, workforces shrink. So, the pool of Business Users that your software must work for is smaller and has different dynamics. Ever larger chunks of the market descend towards platforms made for midmarket. Do yourself a favor and try to write down as many Independent Software Vendors as you can with a portfolio with the breadth of CM.com and with a focus on midmarket. There really isn’t much out there.

Outside of our rich experience in CPaaS, where we also excel with enterprises, CM.com’s products have been midmarket oriented by and large. Midmarket has long been considered no-man’s land. You either sell self-serve to small business (super high volume, low price), or customizable to the enterprise (low volume, high price) if you want to scale a tech company. The stuff in between has long been considered “scraps”.

We saw this change coming and prepared accordingly. We’ve been updating our products to enterprise non-functionals (security and compliance features for example) but have kept our core functionals (business features) focused on midmarket metrics. We already see the success of logo’s flocking towards CM.com that used to be regarded as too upmarket and out of reach for us. We expect this trend to persist, as larger companies get leaner and get to experience the breath of fresh air of user-friendly, beautiful, intuitive software that’s available in this piece of the market where we have already made our mark.

In short, scaling in CPaaS in something we’ve been good at for a long time, but going upmarket there is different from going upmarket in our other product categories. So, though we’ve done well with large enterprises on the CPaaS side of the business, we haven’t gone to that market segment with the other product categories. And now, instead of us venturing there, we see that piece of the market coming to us. As AI accelerates, so will the trend of organizations “right-sizing” to new efficiency standards. We’ll be there to welcome them.

The Conversation as an Interface

A big part of what people refer to when they say user-friendly is of course the interfaces of software tools. We pride ourselves on sweating the details to make software experiences that are nice, simple and intuitive for our clients.

For consumers such a nice, simple and intuitive experience quite often means a conversational one. Talking is a skill most humans master starting at 2 years old. They don’t come to your website or portal every day. They don’t know the nitty gritty of your organization, its policies and processes at all. So, you take them by the hand with marketing, sales, service, etc.

Asking a question and getting a knowledgeable answer and set of actions in return by a skilled person is almost always the fastest way from A to B. AI is getting só good, só capable, however, that it can take on those roles now too. And it’s not just for consumers. Controlling CM.com applications is becoming more conversational too. Fewer buttons, fewer clicks, more chatter.

Natural Language, in the form of text and speech is becoming the default method to interact with AI systems. Instead of us changing our ways to talk to computers, computers change their ways to talk to us. Beyond that even, we’re getting into the territory of cases where they don’t need any instructions at all.

CM.com is an expert on this conversational subject since its founding. From being one of the very first in the world to offer A2P SMS in 1999, to being the most channel-agnostic vendor today. The conversational interface is in our DNA. As truly capable AI takes the stage, our rich history in CPaaS, CCaaS, Conversational AI and MarTech products to orchestrate customer engagement gives us (and by extension our clients) a very complete set of capabilities to create these Conversational Experiences with AI. With our Payments and Ticketing products, we make these experiences transaction capable as well for end-to-end Conversational Commerce.

Take the list of companies with a broad product portfolio and truly solid AI capabilities in the midmarket space you had in mind earlier. Now narrow it down to those with their own channels, payments and other transaction capabilities.

This long term-vision of CM.com has never changed, it’s just taken a while for the world to get into a position where all puzzle pieces start making sense.

The Transition From “Traditional Software” to AI

In 2017, Jensen Huang (NVIDIA) followed up on Andreesen’s essay with the statement:

“Software is eating the World, but AI is going to eat Software”

It was a seminal year for AI, with Alec Radford discovering that “old” models like LSTMs generalize if you make them big enough and a few months after that, the now famous introduction by Google of the model type that has been ruling the world since: Transformers. The starting gun for the race towards AGI and ASI had been fired.

Neural Scaling Laws for Training ánd Test-Time Compute were discovered, new ways to synthesize training data, algorithmic improvements and of course all the best talent and funding in the world is flocking towards AI. This almost magical thing that gets predictively better as you give it more resources…is getting a lot of resources…

If by now, you don’t see that this trend has no ending… That it’s the be all, end all of innovation, we’re afraid you’re asleep at the wheel.

If you’re awake though, keep reading.

You see, we’ve been preparing for the seismic effects AI will have on our clients and the role we feel we need to play to help them. So, what does it mean concretely when we say, “AI is going to eat software”? And why are we so bullish on our software if AI is going to eat it?

There will be, and there are already, loads of upstarts and mature tech companies shooting for Agent Products too. Just like an employee in a firm though, an AI agent does not operate in a vacuum. It needs to work together with other Agents (AI and/or Human) and applications. That’s where the magic is. And for us, our Agentic Products will do exactly that. Working as intelligent needles and threads throughout our entire product portfolio. Stitching together capabilities in depths that you can only achieve when you own all those apps yourself.

Our agents are context-aware and self-enriching by default about everything surrounding your audience. Their profile, conversations, transactions. All stuff they “know” from adjacent CM.com products and need never ask. And they can perform actions deep inside these products. In nooks and crannies that you can only access if you own them, end-to-end. It’s this default awareness, and depth of reach, that makes them capable of end-to-end automation with human intervention getting as close to absolute zero as can be achieved.

During the multi-year transition of the market towards Agent-first software, you must be able to offer the best of both worlds. Software for business users in the classical sense ánd its successor which compliments and gradually overtakes the old up to a certain extent. Companies don’t go from 0 to 100.

That’s why we believe a broad and user-friendly product portfolio is key today, and we already see the newest AI Agent-only upstarts coming to this realization and trying to build platforms like ours too, rather than integrating superficially with every vendor in the market. Logically, all software will have to converge towards similarly broad, user-friendly, Agentic platforms or cease to exist. And our starting point as we see it, is highly favorable.

A frontrunner in AI, focused on the right part of the market, as broad and vertically integrated an Independent Software Vendor as you can find, with the richest conversational history (25+ years) on the market.

There will be those with great AI Agents, but too narrow a scope of products. Those with portfolio’s broad enough, but too slow and ill-equipped to move on AI. And those with broad portfolios and good Agentic capabilities, but a focus on very large company sizes that will grow increasingly uncommon in the future.

CM.com won’t find itself in any of those categories, by design.

Some Final Thoughts

Work in modern service economies went from paper on desks to numbers on screens and that felt like a big change. The same thing happened for the consumer side of things. Today, 9 out of the 10 biggest companies in the world are tech companies, almost anything you use has semiconductors in it, and touching and talking to screens is normal.

At the same time very little changed. As Robert Solow famously put it in 1987:

“You can see the computer age everywhere but in the productivity statistics”.

We can quibble about whether the 2 productivity paradoxes have been a mirage of mismeasurement, but it is common sense that we have not seen a real new industrial revolution – yet. That’s about to change. With AI that can act, not just chat, the digital economy is going to get an infinite labor force, soon. So far, the output of Business Users within software was still a limiting factor on productivity. When all these tools (existing software) start interfacing with super scalable and reliable labor, that cap disappears. We’re ready to really move that needle for our clients.

For the years 2025 and 2026, we make a clear differentiation between different types of agents. Bounded and unbounded agents. One being targeted at creating and executing operational task automations and one at difficult creative tasks, if you will. Fully integrated with our existing products.

The first launch, meant for operational workloads, will already have occurred by the time our annual report comes out. This Agent Studio and Governance App will be expanded upon heavily this year with another key development that we’ll keep under wraps for now. The second category of work will also get a product launch this year.

Beyond that, we’ll enter the third phase of our plan, towards empowering highly autonomous organizations. More about that later, when the time comes. We feel we have our raison d'être figured out. To take our clients with us on this journey, we’re naming our new product family and vista for the years to come: HALO. A circle of bright light around the AI transition. To brighten the path ahead, in a time of change that the world hasn’t seen since the industrial revolution.

Digital conversations will skyrocket. Transactions will too. It will all need to be connected to each other and autonomously operated by AI. And as we’ve seen over the last years, the rise of everything being readily available digitally, synthetically, drives us as people to seek at genuine, authentic, live experiences. CM.com has a striking role to play in all of it.

Our position in the market, the breadth of our portfolio and our know-how with AI are coming together beautifully for us take our rightful place in the world in the years to come.

All in all, we are now more than ever, ready to be at the forefront of innovation once again. Are you?