Ada’s $430M merger, decoded

You are receiving the community edition of today’s newsletter. Become a member to receive exclusive perks and event invites.

On Tuesday, the Globe and Mail reported that Toronto-based Ada was acquired/merged with California’s Dialpad for $430M USD in Dialpad shares. Mike Murchison will join Dialpad as the Chief AI Officer.

While Ada founders have been on the TechTO stage multiple times (see the past presentations link below), we aren’t familiar with the company’s details or the merger. The following is an outside-in analysis.

Why the two companies merged

Both startups are 10+ years old and operate in adjacent markets. Dialpad offers a modern VoIP system that combines business phone services, video conferencing, messaging, and customer service tools into a single workspace. Ada is a customer service platform that helps businesses handle customer inquiries.

Dialpad launched a custom AI model in August 2023, Dialpad GPT, trained on over 5 billion minutes of business conversations. This enabled them to serve their clients with ultra-low latency during live phone calls while maintaining strict data privacy.

As LLMs took off, Ada rebuilt its product to run on third-party models like OpenAI’s GPT-4, moving from scripted chat flows to generative AI. It then launched a ‘Reasoning Engine’ across text and voice and expanded into phone support, image understanding, and task execution through ‘Ada Computer.’

As of this summer, Ada was growing moderately and had $50M USD in ARR, while Dialpad has surpassed $350M ARR. The two companies together serve 50,000 customers.

By any measure, both companies have built large businesses but were facing two strong headwinds:

The rapidly changing AI landscape is putting continuous pressure on the ability to improve the product offering
The requirements to go public keep getting harder, and the market seems to want companies to have a minimum of $500M in revenue and growing fast at 100% YoY

The companies believe that by merging, they can accelerate growth, increase the competitive moat, and lower costs, which should help the combined entity go public sooner.

The unified company can offer smooth handoffs between human contact centre agents and autonomous AI agents and use its massive data layer to automatically update Ada's customer-facing AI agents to resolve incoming problems in real time.

What made Dialpad the acquirer

It has both a faster growth rate and more revenue
It had a higher valuation at the last round: $2.25B vs. $1.2B for Ada
It may have had higher margins as it built its own AI model vs. building on top of other models.

How the deal came together

While the companies didn’t have overlapping board members, they shared investors, including Andreessen Horowitz (a16z), Google Ventures (GV), ICONIQ Capital, and Bessemer Venture Partners, which likely helped with introductions and made the deal smoother.

How Ada’s investors and founders did

This is an outside-in estimate based on assumptions, as we don’t have the details on each round's terms and options distribution as the company expanded.

The exit is valued at $430M USD, a 64% markdown from the last round
Founders start with 10,000,000 common shares at $1.00 each
Each round sells 20% of the company (post-money = 5x the amount raised)
A 10% employee option pool expansion comes out of the pre-money valuation at every round
Every investor holds 1x non-participating preferred shares

The biggest assumption is that every shareholder can sell at $430M. In practice, it isn’t clear if the combined company will go public or at what price. Even if it does, not everyone can liquidate their holdings without crashing the share price.

Exit waterfall summary

If the merger leads to a successful IPO down the road, the deal still creates real value for shareholders, founders and employees, even at a markdown from the last round.

Founders should also know that in today’s market, Series C investors almost always get a 1x liquidation preference. That’s a big reason valuations can be so lofty. The preference protects their downside in a way other shareholders aren’t protected.

If the company does well, investors with a preference share in the upside. If it sells for less than expected, they still get their money back as long as the sale price covers what they put in.

For example, an investor who buys 10% at a $1B valuation puts in $100M. If the company then sells for $100M, they take all $100M and break even, and nobody else gets anything.

The lesson for founders is to know your preference stack as well as your valuation. For a look at where Ada started, here’s the team on the TechTO stage over the years:

🚀 Ottawa week recap, from karaoke to the Rocketships list

Our SAAS NORTH AI week started a night early at Lowertown Brewery, where founders, operators and investors traded the conference agenda for karaoke.

At the conference, we announced Ottawa’s Rocketships and heard from the founders and investors behind them: Backboard.io, Pluvo, DashQ, NuEnergy.ai, miRoncol Health and enurgen. Know a founder who belongs on the next list? Nominate them here →

We wrapped the week over a private dinner with Ottawa founders running $3M+ revenue or venture-backed companies.

Glimpse into TechTO x SAAS NORTH AI week

🎓 “Learning how to use AI is a really good life skill for anything you do in life.” ~ Everest Nevraumont

A three-time Texas State History Champion who spends about two hours a day on core lessons. Everest learns with an AI tutor and adaptive apps, and if you’re a parent, you’re probably already wondering:

What does a kid do with the rest of the day?
Who catches it when the AI gets something wrong?
Does learning that fast still teach a kid to sit with a hard problem?

On Nov 10, Everest and her dad, Edward Nevraumont, join us on stage along with Carly Shuler, co-founder and CEO of Hoot Reading, and Rohan Mahimker, co-founder and co-CEO of Prodigy Education, to take those questions on.