Meta's Enterprise AI Move Knocks 3% Off Salesforce and ServiceNow
Meta's move into enterprise AI software sent Salesforce down 2.88% and ServiceNow down 3.28% on Monday, with MongoDB off 18.33% and Meta itself the biggest faller among the group.
Meta Platforms is pushing into enterprise software with an AI product aimed at territory held by Salesforce and ServiceNow, and on Monday, 28 September 2026, CRM fell 2.88% to 227.27 and NOW fell 3.28% to 131.17 while META itself dropped 5.01% to 714.01, as of 19:57 GMT.
Meta Platforms (META) is pointing its artificial-intelligence effort at enterprise software, the corporate back-office market long dominated by Salesforce (CRM) and ServiceNow (NOW). The move, reported by GuruFocus, introduces a new and extremely well-capitalised competitor into a category that has enjoyed unusually durable pricing power for a decade.
The market did not treat it as a rounding error. As of the last trade at 19:57 GMT on Monday, 28 September 2026, Salesforce was quoted at 227.27, down 2.88% from a prior close of 234.02, and ServiceNow at 131.17, down 3.28% from 135.62. Both fell further than the broad market: the S&P 500 proxy SPY was off 0.80% at $765.21 and the Dow proxy DIA down 0.69% at $513.91.
The whole software complex sold, not just the two named targets
What makes Monday's tape interesting is the breadth of the move across enterprise names, not the depth of any single one. Microsoft (MSFT) traded at 509.26, down 1.34%. Oracle (ORCL) was at 132.76, off 3.17%. Adobe (ADBE) sat at 230.84, down 1.97%, and Accenture (ACN) — the consultancy that implements much of this software for large customers — was at 174.37, down 0.99%.
The outlier was MongoDB (MDB), down 18.33% at 335.19 against a prior close of 410.44, having traded as low as 300.00 during the session. That is a different order of magnitude from everything else on the list and is not explained by sector sentiment alone; a move of that size in a single day is company-specific in character, whatever the surrounding narrative.
The Nasdaq 100 proxy QQQ was down 1.11% at $736.24, so software's losses ran ahead of the tech benchmark. That is the signature of a sector-level rerating rather than a general risk-off day.
Why Meta entering the category matters more than most new entrants
Enterprise software has been structurally hard to attack. Customer relationship management and IT service management systems sit at the centre of how large companies run — they hold the data, the workflows and the compliance records, and ripping them out is expensive and slow. That switching cost is what has allowed incumbents to raise prices steadily and keep renewal rates high.
The argument for why AI changes that is straightforward: if a general-purpose model can read a company's data, reason over it and execute the workflow, then the value that used to sit in a rigid, configured application starts migrating to the model layer. A vendor with frontier-scale compute, its own models and no legacy licence revenue to protect can price aggressively in a way an incumbent cannot without cannibalising itself.
Meta brings all three of those things. It also brings something less obvious: it does not need enterprise software to be profitable. Its advertising business funds the experiment, which means it can run a long campaign at low or negative margin. That is the competitive asymmetry investors in Salesforce and ServiceNow were repricing on Monday.
Meta's own shareholders took the larger hit
The sharpest decline among the megacaps on the list belonged to Meta itself. The shares were quoted at 714.01, down 5.01% from a prior close of 751.66, with a session range of 713.19 to 750.58 — meaning the stock closed the observation window essentially at its low. On a points basis that is a drop of roughly 37.65 from Friday's close.
That is worth sitting with. A company announcing an expansion into a large adjacent market usually gets some benefit of the doubt. Meta did not. The market's read appears to be that enterprise software is a capital-intensive, sales-intensive, slow-burn business with long procurement cycles and demanding customers — the opposite of Meta's consumer advertising model — and that the spending required arrives long before any revenue does.
There is also the broader context of AI capital expenditure. Investors have spent 2026 asking repeatedly whether the largest platforms are spending at a rate their returns can justify. A new business line requiring more infrastructure, more engineers and a direct enterprise sales force reads, to a sceptical shareholder base, as another line item rather than another revenue stream.
What to watch from here
Three things will determine whether Monday's move was a repricing or a reflex.
- Product substance. The distance between a demonstration and a system a Fortune 500 company will trust with customer records and audit trails is measured in years, not quarters. Security certifications, data residency and support commitments decide enterprise deals as often as capability does.
- Incumbent response. Salesforce and ServiceNow both have their own AI agent strategies and both have the distribution Meta lacks. The question for their next reported quarters is whether net revenue retention and average contract value hold, or whether renewals start getting repriced downward in negotiation.
- The services layer. Accenture's relatively modest 0.99% decline is a tell. Consultancies get paid to integrate whatever the customer buys, so they are comparatively indifferent to which vendor wins — unless AI compresses the implementation work itself.
For now the facts are limited: a stated intent, a set of prices, and a market that moved software lower across the board. The revenue at risk for the incumbents cannot be quantified from what has been disclosed, and anyone putting a number on it today is guessing. What can be said is that the market's assumption of permanence in enterprise software pricing is no longer being taken for granted — and that Meta paid a higher share-price cost on Monday for entering the market than either of the companies it is said to be targeting paid for being attacked.
Key facts
- META price: 714.01, -5.01% (as of 19:57 GMT, 28 Sep 2026)
- CRM price: 227.27, -2.88% on the day
- NOW price: 131.17, -3.28% on the day
- Biggest decliner in the group: MDB, 335.19, -18.33%
Frequently asked questions
What did Meta announce?
Meta Platforms is pushing into enterprise software with an artificial-intelligence offering aimed at the market served by Salesforce and ServiceNow. The available reporting establishes the strategic intent and the target competitors; detailed product specifications, pricing and availability timelines have not been set out, so the commercial scale of the effort cannot yet be measured.
How far did Salesforce and ServiceNow fall?
As of the last trade at 19:57 GMT on 28 September 2026, Salesforce was quoted at 227.27, down 2.88% from a prior close of 234.02. ServiceNow was at 131.17, down 3.28% from 135.62. Both fell considerably more than the S&P 500 proxy SPY, which was down 0.80% at $765.21 that session.
Why did Meta shares drop on its own news?
Meta fell 5.01% to 714.01, the steepest decline among the megacap names involved and larger than the drops at either company it is targeting. Enterprise software requires heavy upfront spending on infrastructure and direct sales, with revenue arriving only after long procurement cycles — a cost profile investors already wary of AI capital expenditure treated as a negative.
What happened to MongoDB?
MongoDB fell 18.33% to 335.19 from a prior close of 410.44, trading as low as 300.00 during the session. That decline is an order of magnitude larger than the rest of the software complex and is company-specific in character rather than explained by broad sector sentiment around Meta's entry into enterprise applications.
Why is enterprise software considered hard to disrupt?
Customer relationship management and IT service management systems hold a company's core data, workflows and compliance records. Replacing them is slow and expensive, which creates high switching costs and lets incumbents raise prices and retain customers. The AI argument is that if a model can read the data and execute the workflow directly, that lock-in weakens.
How did the broader market trade that day?
It was a weak session across the board. The S&P 500 proxy SPY closed the observation window at $765.21, down 0.80%. The Nasdaq 100 proxy QQQ was at $736.24, down 1.11%, and the Dow proxy DIA at $513.91, down 0.69%. Software names generally fell further than those benchmarks.
