LivePerson 2025: Cost Cuts Bought Time, but Customer Trust and the Debt Story Still Run the Show

Getting to Know the Company

LivePerson is not just another software company selling ordinary licenses, and it is not just another AI story that can be read through product headlines. It is a SaaS platform for digital conversations between brands and customers, powering nearly one billion interactions a month across websites, apps, SMS, social channels and third-party messaging platforms. Its real economic engine is not the existence of a product, but the ability to persuade large enterprises to renew contracts, expand usage and trust the company with sensitive service and sales workflows.

What is working right now is visible. The company cut costs aggressively, reduced headcount, narrowed operating loss to $78.7 million from $183.2 million in 2024, and lifted Average Annual Revenue Per Enterprise and Mid-market Customer, or ARPC, to $680,000 from $625,000. That means at least part of the surviving customer base is larger or spending more.

But this is still not a clean return to healthy growth. Revenue fell another 22% to $243.7 million, retention dropped to 78% from 82%, far below the long-range target of 105% to 115%, and the company itself says renewals and new bookings are running slower than expected, partly because customers still worry about LivePerson’s financial stability and partly because enterprise buying cycles for high-value AI solutions have lengthened due to compliance and approval requirements. That is the active bottleneck now. Not product relevance, not a missing feature, but commercial trust.

Anyone looking only at ARPC, the cost cuts or the AI narrative could read 2025 as a turnaround. That is too shallow a reading. The better reading is a smaller, leaner company that still has not proved the commercial base has stopped shrinking. A higher ARPC can also come from a smaller business if weaker customers disappear and the remaining base skews larger.

There is also a practical actionability constraint that belongs early in the screen. In October 2025 the company executed a 1-for-15 reverse stock split to regain compliance with Nasdaq’s minimum bid price requirement. If it falls back into non-compliance before October 13, 2026, it will not be eligible for another automatic 180-day cure period. At the same time, the Tel Aviv line is very illiquid, with daily turnover of only about NIS 24.8 thousand on the latest observed trading day. That matters. Even if the business stabilizes operationally, the equity layer still carries listing, liquidity and capital-structure friction.

A compact economic map:

The next table shows why one glance at ARPC is not enough:

Events and Triggers

The first trigger: in September 2025 LivePerson launched a new restructuring plan to align cash spending with current commercial performance. The charge was $11.7 million in 2025, and the company says the program was substantially completed by year-end. That changed the cost base quickly, but it also means a meaningful part of the easy cost reset has already been harvested.

The second trigger: on September 12, 2025 the company completed its large debt exchange. It exchanged $341.1 million of 2026 Notes for $45.0 million of cash, $115.0 million of second-lien notes, 3,555,596 common shares and 26,551 shares of Series B preferred stock. Then, in October 2025, the preferred converted automatically into another 1,547,840 common shares. In plain English: the company bought time, but it paid for that time with both cash and dilution.

The third trigger: the equity layer itself went through capital surgery. The common share count rose to 12.04 million from 6.08 million, after retroactive adjustment for the reverse split. So even after part of the old debt was exchanged away, shareholders now own a smaller slice of a company that still carries a heavy capital structure.

The fourth trigger: the 1-for-15 reverse split in October 2025 restored Nasdaq compliance on the minimum bid price rule, but it did not solve the issue. It only delayed it. If the stock falls back out of compliance before October 2026, the company will not get another automatic cure period. Because a Nasdaq listing failure is also a fundamental change under the 2026, 2029 and second-lien note indentures, this is not just a stock-market optics issue. It is a capital-structure issue.

The fifth trigger: in January 2026 the board assigned Nathan “Tripp” Lane to the Audit Committee and Ryan L. Vardeman to the Compensation Committee, after their appointments as directors in October and November 2025. This is not an operating trigger by itself, but it does suggest that part of the 2026 agenda will run through capital discipline, oversight and rebuilding trust.

The meaning of that chart is double-edged. Management did move the cost base. But after a cut of this size, the question is no longer whether the company can trim more. It is whether it can stop the revenue slide without further damaging its sales, service and product capacity.

Efficiency, Profitability and Competition

The central story of 2025 is a paradox: operating loss improved sharply, but the core unit economics of the business weakened. Revenue fell 22%, while cost of revenue fell only 10%. That means cost of revenue rose to 28% of sales from 25% in 2024, so gross margin compressed. This matters. The efficiency work happened below gross profit, but it did not stop erosion in the revenue engine itself.

The company cut $21.5 million from sales and marketing, $35.3 million from general and administrative expense, and $25.1 million from product development. Much of that came from lower salary, stock-based compensation and other employee-related costs tied to restructuring. So the improvement in operating loss to $78.7 million does not mean LivePerson is back to efficient growth. It means the company cut fast enough to keep the revenue decline from damaging the bottom line at the same pace.

The most interesting detail in the filing is that the commercial deterioration does not look like the loss of one anchor customer. The company explicitly says no single customer accounted for 10% or more of revenue in 2025, 2024 or 2023. So this is not a hidden concentration story. It is a broader erosion story: cancellations, smaller renewals and slower new business formation.

That chart matters because it breaks the decline into layers. Revenue in the Americas fell hard, from $219.3 million to $134.4 million, while EMEA and APAC actually grew. That does not mean the problem is simply geographic. But it does show that the heaviest pressure sits in the largest and more mature commercial base. If management manages to turn the business, that turn will probably have to begin with stabilizing the Americas.

Another layer that looks better on first read than on second read is the contracted base. Remaining performance obligations stood at $175.6 million at the end of 2025, and roughly 98% is expected to be recognized within 24 months. That gives some visibility, but not enough to settle the growth question. When annual revenue is $243.7 million and current deferred revenue fell to $54.3 million, the commercial issue is clearly not being solved by a hidden reservoir of already-signed business sitting outside the income statement.

On the competitive side, LivePerson is not only fighting legacy names such as Genesys, Oracle, Salesforce and Twilio. It is also facing a new wave of AI-native competitors. The filing says the environment is defined by aggressive investment from competitors with significant resources, and that enterprise AI deployments often require additional compliance and approval cycles. That is the difference between an impressive demo and recognized revenue. In this market, technology by itself is not enough. Trust, time and a reassuring balance sheet also matter.

Cash Flow, Debt and Capital Structure

The right cash frame

Here the right frame is all-in cash flexibility, not a normalized cash-generation frame. The reason is simple: the LivePerson thesis now sits on financing flexibility, margin for error and capital structure, so the relevant question is how much cash is really left after actual cash uses, not how much the business might generate before those commitments.

On that basis, 2025 was weak. Cash ended the year at $95.0 million, down from $183.2 million. Operating cash flow was negative $30.4 million, investing used another $13.7 million, and financing used $45.5 million, almost all of it driven by the $45.0 million paid to lenders in the debt exchange. This is not the picture of a business generating its own room for error.

Management does say current cash should be sufficient for at least the next 12 months, but that needs to be read next to the hard facts: the 2026 Notes still require $20.1 million of repayment in December 2026, and the 2029 Notes require the company to maintain a minimum cash balance of $60.0 million at all times. That means the real margin for error is not wide. It would be wrong to say only $15 million is literally free today, because cash flow can still change during the year, but it is correct to say the cushion above the covenant floor and the near-term maturity is narrow.

The debt exchange: a real improvement, but not a clean one

The positive side is obvious. At the end of 2024, debt carried on the balance sheet stood at $527.1 million. By the end of 2025, net carrying debt had fallen to $391.8 million, and the current portion had shrunk to only $20.1 million. The stockholders’ deficit also improved to $44.5 million negative from $67.3 million negative.

But the accounting numbers are misleading if you stop there. The second-lien notes were recognized on the balance sheet at $182.0 million even though their stated principal is only $115.0 million. The reason is troubled-debt-restructuring accounting, which loads the liability with maximum future interest and a redemption premium. That creates an accounting gain at the exchange date, but it does not change the fact that the credit market is still pricing this layer very conservatively.

That chart may be the most important one in the entire filing. The 2026 Notes carry at $20.1 million, but fair value was only $8.2 million. The second-lien notes carry at $182.0 million, but fair value was only $61.5 million. Even if the 2029 Notes trade closer to economic value, the overall credit stack still does not look like a fully repaired capital structure. Put differently, the balance sheet calmed faster than the credit market did.

The price of capital is also still far from normal. The effective rate on the 2029 Notes at year-end 2025 was 19.13% for the original 2029 Notes and 13.28% for the Delayed Draw Notes. The second-lien notes carry a 10.0% coupon, initially as PIK, and management valued them using a 30.0% yield assumption. These are not the numbers of a business that has returned to ordinary financing conditions.

Dilution, deficit and what really belongs to equity

For common shareholders, the cost is very clear. Common shares outstanding rose to 12,039,325 from 6,079,378, almost a doubling. At the same time, stockholders’ equity remained negative by $44.5 million, and accumulated deficit widened to $1.058 billion. So it is not correct to read the debt exchange as a story where “the debt went away.” Part of it was pushed out, part of it was converted, and part of it simply changed layers.

There is another real claim on flexibility as well. In September 2025 the company entered a new three-year IT infrastructure and cloud commitment totaling $74.4 million, with $25.5 million due in 2026, $24.5 million in 2027 and $24.4 million in 2028. That may be strategically sensible as part of the public-cloud migration, but it also reminds you that flexibility is constrained not only by debt, but also by operating commitments.

Forward View

Four points matter before looking at 2026:

• LivePerson is entering 2026 as a proof year, not a breakout year.
• A rising ARPC does not settle anything while retention, deferred revenue and contract acquisition costs continue to fall.
• The debt exchange bought time, but it did not erase the weak equity layer or the high cost of capital.
• The market will now measure less of the AI narrative itself and more of whether customers are actually willing to renew and expand.

The challenge is that the filing does not provide a single hard revenue target for 2026 that can settle the debate in one number. Management does, however, provide a clear direction: current cash should be sufficient for 12 months, but additional capital may still be needed; renewals and new bookings are slower than expected; and the company is still investing in cloud migration and the platform. So the question is not “what is the guidance?” but “what kind of year is this?” The answer is a commercial and financing proof year.

What could the market miss on first read? First, the contraction is not uniform, and EMEA and APAC still grew. That means there are still working parts inside LivePerson. Second, because no single customer exceeds 10% of revenue, a recovery does not depend on one giant logo, but it also means the problem itself is diffuse, so there is no quick one-deal fix. Third, a large share of the 2025 improvement already came from cost cuts. That means the next report will have to show more revenue evidence and less cost-reset evidence.

What has to happen over the next 2 to 4 quarters for the thesis to strengthen? First, retention at least has to stop falling and start moving back toward management’s long-range target. Second, bookings and renewals need to show that customer concern around financial stability is easing. Third, the cash layer needs to stop eroding at the same pace, because the distance from the $60 million covenant floor is not large. And fourth, the cloud migration and infrastructure spending need to translate into a better commercial proposition, not just a higher fixed-cost base.

What could break the thesis? If renewals continue to close at smaller sizes, if AI growth stays mostly at the narrative level rather than becoming signed revenue, and if the Nasdaq question returns by mid-2026 without clear commercial improvement, the market could easily reinterpret all of 2025 as another year of buying time rather than a true inflection.

Risks

The first risk is continued erosion of the customer base. A 78% retention rate is not a cosmetic metric. It is evidence that the base is still shrinking even after upsells, downsells and churn. As long as that number does not turn, a higher ARPC may simply describe a smaller business with fewer, larger customers.

The second risk is capital structure and continued listing. Another failure on Nasdaq’s minimum bid price rule before October 2026 would not receive another automatic cure window, and delisting would count as a fundamental change under all major note indentures. In other words, the equity and debt stories are no longer separable.

The third risk is hard commitments against a smaller revenue base. A $74.4 million three-year IT and cloud commitment can be read as the right investment, but if revenue keeps falling it can also become another layer of pressure on flexibility.

The fourth risk is further impairment risk. In 2025 the company recorded another $41.6 million goodwill impairment, and even after that charge goodwill still stood at $184.9 million. When the equity trades under stress, the business is still shrinking, and the filing itself shows sensitive valuation assumptions, it is hard to treat that goodwill balance as a comfortable cushion.

The fifth risk is the legal and governance noise that still sits in the background. The U.S. Damri securities case was dismissed and later affirmed on appeal, but remanded with leave to attempt re-pleading, while the parallel Tel Aviv litigation remains stayed. This is not the core thesis today, but it is another layer of distraction and possible cost.

Conclusions

LivePerson ended 2025 in a better place than where it started the year. Costs were cut sharply, operating loss narrowed, and the debt exchange pushed the immediate wall further out. But this is still not a healthy software-company reset. The customer base is still eroding, the contracted layer is not rebuilding convincingly, and the credit market is still pricing the capital structure with heavy skepticism.

Current thesis in one line: LivePerson bought time through cost cuts and debt exchange, but the real test has now shifted to rebuilding customer trust and stopping the capital structure from taking over the story again.

What changed from the older reading of the company is fairly clear. The question is no longer only whether it survives the old debt stack. It is whether, after all the cuts and exchanges, there is still a business here that can grow again without leaning on dilution, another reverse split or another financing round.

The strongest counter thesis is that the restructuring has already done most of the work it can do, while the growth engine has still not returned. If that is right, 2025 will look in hindsight like a technical stabilization year rather than a real turning point.

What could change the market’s near- and medium-term reading is not another AI headline, but very concrete numbers: better retention, a stop to the decline in deferred revenue, healthier renewal sizes, and a cash balance that does not erode too quickly relative to the covenant floor.

One sentence on why this matters: in LivePerson’s 2025 setup, the question is no longer whether there is a product, but whether there is enough trust to convert product relevance into cash and prevent debt from dominating the story again.

What has to happen over the next 2 to 4 quarters is also fairly clear. The company needs to show real improvement in retention, healthier renewals from existing customers, evidence that it can win new business despite elongated buying cycles, and cash stability against the $60 million covenant floor. What would weaken the thesis is a combination of continued revenue erosion, renewed share-price pressure on Nasdaq, and a need for additional financing before the commercial base has actually turned.