Can Palantir’s slowing growth still support a Buy rating? Goldman Sachs focuses on upward revisions and repeatable delivery

Why might growth still exceed expectations as it slows?

Goldman expects Palantir’s growth to slow over the next 12 months, but sees room for earnings expectations to be revised upward. Its model forecasts total revenue growth falling from 87.0% in 2026 to 58.1% in 2027 and 46.1% in 2028; even after the slowdown, growth in 2027 would remain well above the roughly 25% level of high-growth software peers.
The U.S. commercial business illustrates the distinction: revenue rose 150% year on year in the second quarter of 2026, while market consensus implies approximately 137% growth in the third quarter, against a higher year-earlier base. Goldman considers third-quarter growth above 145% a reasonable upside scenario, corresponding to revenue for the business approximately 3% above consensus; beating expectations does not require growth to exceed the second-quarter rate.
Stronger growth also has a clear threshold: Goldman’s optimistic scenario assumes the U.S. business grows by more than 100% for each of the next six quarters. Whether it can sustain high growth will determine the extent of upward revisions to expectations.

Goldman argues that valuation should be considered alongside growth: enterprise value to free cash flow divided by the growth rate is approximately 0.9 times for Palantir and 2.7 times for peers; but on a PEG basis using GAAP earnings, both are approximately 2.1 times, so the discount does not appear under every measure.
The absolute valuation remains high: Goldman’s table puts the forecast price-to-earnings ratio for 2026 at 122 times. The bank maintains its Buy rating and 12-month price target of 230 dollars, 17.2% above the share price of 196.19 dollars; the target applies a multiple of 60 times to 2029 GAAP earnings per share, then discounts it back to the forward period used in its coverage framework. The thesis therefore rests on sustained high growth and upward earnings revisions, rather than a low multiple alone.

Microsoft, Snowflake, Databricks and frontier model companies are building teams of forward-deployed engineers. Goldman believes Palantir’s advantage lies in feeding customer deployment experience back into its products, business relationship modeling and industry solutions, then turning knowledge from individual projects into reusable software; replicating the team’s headcount is only one part of the equation.
Whether this thesis holds depends on whether AI deployment engineers and AIP Evolve can incorporate more experience into the product, making business modeling, process configuration and application deployment faster and more automated. If growth continues to depend on traditional manual deployments, delivery capacity could constrain expansion.

Goldman acknowledges that, after a small number of high-return projects, continued growth requires constantly finding new uses. Its estimate caps a single customer’s budget for Palantir at 10% of software and services spending, corresponding to a serviceable market of approximately 50 billion dollars and approximately 35 million dollars per company among the world’s top 2000 enterprises; current annualized average revenue per commercial customer is approximately 4 million dollars.
The path to expanding the opportunity lies in industry-specific solutions and deeper operational use: turning deployments at benchmark customers into repeatable solutions can help win customers in the same industry; Goldman expects AI to move further into operational processes, decision systems and industry applications in 2027. Growth consequently depends on deploying new use cases and expanding customer contracts; slower enterprise AI adoption would directly weaken this thesis.