After closing at $7.85 on Tuesday, SoundHound AI recovered to $8.22 on Wednesday, a nearly 5% increase due to news that was either a masterstroke or a mess, depending on who you asked. The market took about twenty-four hours to decide whether to celebrate or panic after the company announced that it was purchasing LivePerson in an all-stock deal with an enterprise value of about $250 million. It was primarily a celebration. But not before selling first, which seems to be the current strategy used by tech investors: act, then consider.
A specific kind of story is told by the numbers surrounding SoundHound. In Q4, revenue increased 59% year over year to $55 million. That’s not a story disguised as a growth rate; it’s a real one. However, the stock is currently about 60% below its October high of $22.17 and has dropped more than 20% this year. The debate over whether SoundHound is a voice-AI specialist that surreptitiously wins enterprise contracts or a speculative story that got ahead of itself during the 2025 AI frenzy is somewhere in that gap.
The question is sharper because of the LivePerson deal. Instead of using cash, SoundHound is using its own stock, which is priced against a 10-day volume-weighted average with a $7–$12 per share collar. The detail that alarmed people on Tuesday was that floor at $7. More shares are printed to close the gap if SOUN declines, giving current shareholders smaller portions of the market. Additionally, as part of the deal, SoundHound is restructuring LivePerson’s $261 million in secured notes. It is one thing to take on the debt of another business. Another is to absorb it along with the operational difficulties that accompany a problematic acquisition target. How smoothly this integrates is still unknown.
However, if you look closely, the strategic reasoning is not absurd. Twelve of the top 15 international banks and twenty-five of the Fortune 100 are served by LivePerson’s conversational cloud platform. SoundHound offers its in-house voice stack, which is what really functions in smart devices, drive-thrus, and automobiles. When combined, management projects revenue of $350 to $400 million in 2027—possibly $500 million with cross-selling. For a business that currently makes about $55 million a quarter, those are significant figures. Investors appear to think that part of the forecast is accurate. Probably not all of it. Enough.
Following the announcement, D.A. Davidson reaffirmed its Buy rating at $14, recognizing integration risk while characterizing the deal as accretive on a valuation basis. That cautious optimism seems about right. In the meantime, sentiment has remained in what one platform refers to as “extremely bullish” territory on retail forums like Reddit and Stocktwits. Retail holders believe that SoundHound is one of those AI names that the institutions will eventually need to return to. According to reports, T. Rowe Price increased its position by more than 2,000% in a recent quarter. If this is true, patient holders would use this information to convince themselves not to sell.
It’s difficult to ignore how much of the SOUN story relies on execution rather than concepts as you watch this play out. The technology is functional. The list of clients is authentic. Revenue is increasing rapidly. However, profitability is still elusive, the number of shares continues to rise, and the company recently faced its largest integration challenge to date. There is a skeptic who looks at the dilution math and the negative margins and sees a long road ahead for every believer who believes that $20 or more is inevitable once the numbers click. SoundHound truly exists somewhere in between those two perspectives, closer to intriguing than secure and closer to promising than proven. More will be revealed by May 7 and the Q1 print than by any press release.
