Prediction: SoundHound AI Stock Could Be Worth $20 in 2030. Here’s the Math

SoundHound AI’s Q2 revenue rose 45% YoY to $62M, and the stock popped more than 10% after its Aug. 5 results as the company raised full-year guidance. Management cited growing traction for Oasys, including an eight-figure commitment signed in under 90 days from demo to contract, supporting 2026 revenue expectations of $245M (midpoint), up 45% YoY. The article also flags a potential growth tailwind from its planned acquisition of LivePerson (expected to close by end-2026) to expand into cloud communications.

Analysis

SOUN is trading less like a software company and more like a proof-of-product story, which is dangerous if the next leg of growth is still heavy on pilots and customer enthusiasm rather than repeatable expansion. The market should care more about conversion efficiency than TAM size: if each new deployment requires meaningful services support, the revenue ramp can stay strong while gross margin and free cash flow lag, forcing multiple compression even with good topline prints.

The second-order winner is probably not the obvious chip supplier but the broader contact-center and enterprise software stack that gets forced to defend budget share. That creates pressure on incumbents to bundle voice and agentic features faster, which can cap pricing power across the category; LPSN is the most direct public read-through, but the more important dynamic is that larger platforms can commoditize the feature set before SOUN scales enough to matter.

The key risk over the next 1-3 months is that the stock has already discounted a clean acceleration story, so any evidence of lumpiness in bookings, margin dilution, or slower enterprise conversion can hit hard. Over 6-18 months, the falsifier is simple: if revenue growth slides below the high-30s while operating losses remain sticky, the market will stop paying a narrative premium and re-rate the name toward lower-growth software multiples. Conversely, if Oasys keeps converting quickly and LivePerson closes without dilution, the setup can extend, but that is a higher bar than the current price implies.

Consensus is probably right that the category is growing, but too optimistic on how much of that growth SOUN can capture without larger vendors absorbing the use case. The contrarian view is that the current setup is a quality-vs.-story trade: quality software names with clearer margins may outperform even if SOUN keeps posting fast revenue growth, because investors will eventually demand evidence that growth is durable and profitable, not just headline-grabbing.

Key Decisions for Investors

• Do not chase the post-earnings strength in SOUN; wait for a 10-15% pullback or a failed breakout before establishing risk. If entered, prefer a 3-6 month call spread over stock to cap downside from a de-rating.
• Relative-value idea: long SOUN / short LPSN only as a small tactical pair if LPSN trades on takeover hope without deal certainty. The trade works if SOUN keeps taking share and standalone conversational AI names get marked down; exit if a formal LivePerson deal is announced.
• Set a hard alert on the next quarterly guide: if SOUN growth drops below ~35% or operating loss fails to improve, fade the stock with short-dated puts or reduce exposure. That is the point where the market will likely stop paying for the narrative.
• No standalone NVDA trade on this story; any compute read-through is too indirect to underwrite. Use NVDA only as a hedge if broader AI software multiple compression starts to spill into the chip complex.