recent

Download: Insights for success in AI-driven organizations

Ideas for innovation from MIT Sloan Management Review

4 ways the US election could impact 2025 climate policy

Credit: Rodion Kutsaiev / Unsplash

Ideas Made to Matter

Entrepreneurship

How AI helps acquired businesses grow

By

So you just acquired a small- or medium-sized business. You’ve done your research, stayed ahead of competing offers, and carefully threaded the needle to close the deal.

Now the real work begins: adding value to your acquisition.

During the recent MIT Entrepreneurship Through Acquisition Summit, a panel of successful entrepreneurs and investors shared how they’re growing companies and portfolios with the help of artificial intelligence. Here’s a closer look at three takeaways from the panelists.

When vetting AI tools, think like a hedge fund manager, not a VC

As a former venture capitalist, Mark Valdez knows the temptation of a new technology’s buzz, or the lure of a software company’s rocketing stock prices. But as the current CEO of Eads Bridge Holdings — which acquires and tech-enables traditional, nontechnical businesses — he’s mindful not to take too many risks. That’s why, when it comes to vetting AI tools to use with acquired companies, he recommends taking a hedge fund manager approach rather than one of a betting venture capitalist.

A good first step for this method is to evaluate the health of the company offering the AI tools.

“Understand what could go wrong, because even big companies that raise lots of money, do lots of press, do all the right things — they go out of business,” Valdez said. “You want to be thoughtful about the new companies that are coming out.”

He also advised speaking with existing customers and asking hard questions during sales calls with potential vendors. “Understand how much capital the company has raised, who are the investors, what’s their business model, what do they think about the market, and what’s the road map,” Valdez said.

Once you’ve decided on a tool to use, consider experimenting with it in small ways that don’t risk the whole company on a result.

Choose the right work for AI augmentation

Scott Barstow, a technology and software operating partner at Pacific Lake Partners, said that when it comes to thinking about which work can be improved with AI, he considers four areas:

  1. Operational and functional areas of the business. The company’s functional leaders tend to be responsible for researching and suggesting interesting ways to use AI technology to solve problems.
  2. Ways to drive additional customer value through AI-enhanced products and services. This work relies on AI to make incremental improvements. For example, Barstow works with a company that manages job descriptions for large organizations. Last year, the company integrated OpenAI’s technology to allow its customers to augment generating job descriptions based on other job descriptions OpenAI collects and uses to train its models.
  3. Novel products, often based on unique data already owned by the company. Using the same job description example, this next step might involve tapping the company’s database of descriptions and creating a new product using proprietary information that existing customers or even new ones might use.
  4. AI-first companies. Barstow said that this is not typical in the entrepreneurship-through-acquisition space, as the ETA model tends to revolve around buying businesses that are 10 – 15 years old rather than young startups with AI at their core.

Start small, and include employee input

While you might be ready to fully commit to creating value with AI, don’t make the mistake of taking an all-or-nothing approach, said Linda Drabik, board adviser and founder at Marencik Solutions. Instead, the AI and machine learning consultant suggests that leaders start with several clearly defined use cases.

“Deploying a proof of concept or pilot can save time and money in the long run because they focus leadership teams on making better use of data, fast-track a product or feature to improve customer experience or an internal process, and clearly measure value to drive decision-making for downstream tech investments,” Drabik said.

She also advised that entrepreneurs avoid buying into hype or being swept up in flavor-of-the-week tools, a view shared by Valdez. “You don’t have to be the first adopter, but you don’t want to be the last one either,” he said. “Finding that point in time where it matches up for you is really valuable.”

Drabik said employees often fear that AI will replace their roles, so it’s important for leadership to openly discuss the technology’s role in the company’s evolution. Emphasize that working alongside AI can be a way to upskill a team while leaning on members’ expertise. This can improve employee satisfaction and retention post-acquisition.

Various colorful lightbulbs

Entrepreneurship Development Program

In person at MIT Sloan

Drabik also recommended exploring opportunities for using foundation models to quickly integrate AI into existing systems and workflows. And, she said, don’t forget that data governance and data quality are essential to success with AI — things that running a comprehensive data audit can address.

“In addition to offering a view of existing data assets, an audit can highlight opportunities to capture or transform more actionable data and insights, revealing the lift a new CEO can bring to an organization by responsibly applying AI,” Drabik said.

Choose use cases that deliver company-specific value

At intelligence process automation company Beecker, engineers use AI to audit software code.

When you have a coding issue in a small company, everyone can gather around a computer to try to solve the problem together, said Felipe Corcuera, MBA ’18, a board member and co-CEO of Beecker. But that’s not possible for a company with 250 people spread out across the U.S., Colombia, and Mexico. So instead, Beecker uses ChatGPT as a digital coworker to answer questions about why a script is failing or an error message keeps popping up on a screen, for instance.

“It’s almost like having a big brother next to you saying, ‘Hey, you should look at this or that, or make sure this variable is well defined in this other piece,’” Corcuera said.

This also helps employees triage issues, Corcuera said. Instead of going to the chief technology officer with a coding error, a junior employee can run the code through ChatGPT first, and if the problem persists, they can then escalate the issue.

One of the first ways editing and proofreading service Scribendi used AI was for taking customer orders.

“We thought that the business had solid fundamentals and [that] AI could create additional opportunities,” said Enrico Magnani, a founding partner at private equity firm Magnum Capital Partners, which acquired Scribendi in 2015. “I like to think about AI as a bicycle for your mind. Let’s say your brain used to walk, and then you hop on a bicycle: A whole new world of opportunities opens up.”

A few years after Scribendi’s acquisition, the company added a grammar correction tool trained to support editors of scientific articles, another tool to help clients choose the appropriate product, and a third tool for ensuring that orders were fulfilled on time. The company was ultimately acquired by Japanese editing and researching company Edanz in 2023.

For more info Meredith Somers