Most companies haven’t yet seen financial returns from their AI investments, according to PwC’s 29th Global CEO Survey.
The survey of 4,454 chief executives across 95 countries found that 56% report neither increased revenue nor lower costs from AI over the past 12 months.
About 30% of CEOs said their company saw increased revenue from AI in the last year. On costs, 26% reported decreases while 22% said costs went up. PwC defined “increase” and “decrease” as changes of 2% or more.
Only 12% of companies achieved both revenue gains and cost reductions. PwC called this group the “vanguard” and noted they had stronger AI foundations in place, including defined roadmaps and technology environments built for integration.
For marketing specifically, the numbers suggest early-stage adoption. Just 22% of CEOs said their organization applies AI to demand generation to a large or very large extent. The company’s products, services, and experiences showed similar numbers at 19%.
Parking_Act3189 on
How convenient that a PwC study concludes that companies should continue to use PwC and not do that work in much cheaper ways with AI firms or internally.
chrisni66 on
One of the problems I’ve been seeing in my industry is that senior execs will say ‘we need to roll out AI’ without actually stating what they want to use it for. Any technology roll out needs to be predicated on the use case(s) otherwise they are doomed to fail from the start.
There was an analyst study that went on to more detail on this and found that the success rate of AI roll out increased dramatically when it was led not by the desire for AI, but because it was the best tool to solve a problem.
TraditionalBackspace on
My company has spent hundreds of thousands on training and hype. Literally nothing has come from it. Not even simple searches of internal standards, which is one of the only use cases for us at the moment. They are so good at stepping over dollars to pick up pennies.
mywifesoldestchild on
I’d wager a good chunk of the 44% aren“t seeing a revenue gain from AI either. I’d be wary of the honesty in asking a CEO, „Are you having success in your company’s primary initiative?“
costafilh0 on
44% of CEOs report revenue gains from AI?
**AMAZING**
WesternFungi on
Half the time I tell it my software version my problem and it spits out an answer I try it out doesn’t work, repeat the question and the AI goes „oh this command doesn’t exist in this version“ so its genuinely a waste of time I’d rather spend time reading the helpguides and wikis.
RandeKnight on
Well, yes, DUH!
AI is used to **reduce costs**, not increase revenue. The only way it increases revenue is if they are selling AI.
Malkovtheclown on
I’ve noticed a quiet shift to hiring more people on the implementing side of things for AI products. I think everyone is realizing finally Agents aren’t actually plug and play. Often companies are buying Agents to replace existing process when in reality Agents are a brand new way of doing business and require process transformation to be successful.
mightsdiadem on
I dont think anybody understands what this is good at doing yet.
I hear it being used for emails and writing up reports and some programming, but the programming always needs massaged to get it to function properly.
I tried to train it to advise on the system I work on, but my first attempts haven’t panned out. I was thinking I could train it to advise on issues for how I set up the system(s) but I keep getting different results over time, which is incredibly frustrating.
In other words, I train it and get it where it’s answering questions accurately enough to be useful. Then about a week later it’s like I have to start over.
Zerocordeiro on
At work we’re building a dashboard and it’s kind of a first for most of us. My colleagues let the copilot assistant do almost all the coding, then they have no idea how the system actually works. When time comes to do maintenance they can’t evaluate whatever may be going on and almost nothing’s planned as reusable components. It’s a loss for the company mid and long term
Illlogik1 on
That means THEY (the CEOs) are still making shit decisions even though AI is providing quicker , more accurate, and higher quality information to them and their staff, or their company is so bankrupt on creativity and reliant upon their crappy inefficient corporate machine they made to realize the value and profit
dranaei on
AI is a long term investment. You have to be an idiot to expect profit short term. Especially when WE DON’T HAVE CAPABLE MODELS YET.
Subtifuge on
Ironically a bunch of companies forcing LLMs into their products, making their profitable companies become unprofitable and less userfriendly resulted in me using LLMs to just build self hosted replacement products for a bunch of subscription orientated web base apps, I imagine I am one of many, so not only are they losing money though their poor investment stratergies and ramming things people do not want down our throats, they are likely also losing money from competetor „legacy“ products that people are building to replace them,
itsTF on
44% seeing gains is pretty incredible! Lots of companies are attempting to roll out nonsensical AI features, etc. I would’ve expected only 10-20% would actually be doing it in smart enough ways to see benefits from it.
dai_quangling on
Check out for this project creation and deliverables based on user needs first three clients are free and there is 3 major categories for development of projects. Check out: [https://zyro-dev.vercel.app/](https://zyro-dev.vercel.app/)
Leave A Reply
Du musst angemeldet sein, um einen Kommentar abzugeben.
16 Kommentare
Most companies haven’t yet seen financial returns from their AI investments, according to PwC’s 29th Global CEO Survey.
The survey of 4,454 chief executives across 95 countries found that 56% report neither increased revenue nor lower costs from AI over the past 12 months.
About 30% of CEOs said their company saw increased revenue from AI in the last year. On costs, 26% reported decreases while 22% said costs went up. PwC defined “increase” and “decrease” as changes of 2% or more.
Only 12% of companies achieved both revenue gains and cost reductions. PwC called this group the “vanguard” and noted they had stronger AI foundations in place, including defined roadmaps and technology environments built for integration.
For marketing specifically, the numbers suggest early-stage adoption. Just 22% of CEOs said their organization applies AI to demand generation to a large or very large extent. The company’s products, services, and experiences showed similar numbers at 19%.
How convenient that a PwC study concludes that companies should continue to use PwC and not do that work in much cheaper ways with AI firms or internally.
One of the problems I’ve been seeing in my industry is that senior execs will say ‘we need to roll out AI’ without actually stating what they want to use it for. Any technology roll out needs to be predicated on the use case(s) otherwise they are doomed to fail from the start.
There was an analyst study that went on to more detail on this and found that the success rate of AI roll out increased dramatically when it was led not by the desire for AI, but because it was the best tool to solve a problem.
My company has spent hundreds of thousands on training and hype. Literally nothing has come from it. Not even simple searches of internal standards, which is one of the only use cases for us at the moment. They are so good at stepping over dollars to pick up pennies.
I’d wager a good chunk of the 44% aren“t seeing a revenue gain from AI either. I’d be wary of the honesty in asking a CEO, „Are you having success in your company’s primary initiative?“
44% of CEOs report revenue gains from AI?
**AMAZING**
Half the time I tell it my software version my problem and it spits out an answer I try it out doesn’t work, repeat the question and the AI goes „oh this command doesn’t exist in this version“ so its genuinely a waste of time I’d rather spend time reading the helpguides and wikis.
Well, yes, DUH!
AI is used to **reduce costs**, not increase revenue. The only way it increases revenue is if they are selling AI.
I’ve noticed a quiet shift to hiring more people on the implementing side of things for AI products. I think everyone is realizing finally Agents aren’t actually plug and play. Often companies are buying Agents to replace existing process when in reality Agents are a brand new way of doing business and require process transformation to be successful.
I dont think anybody understands what this is good at doing yet.
I hear it being used for emails and writing up reports and some programming, but the programming always needs massaged to get it to function properly.
I tried to train it to advise on the system I work on, but my first attempts haven’t panned out. I was thinking I could train it to advise on issues for how I set up the system(s) but I keep getting different results over time, which is incredibly frustrating.
In other words, I train it and get it where it’s answering questions accurately enough to be useful. Then about a week later it’s like I have to start over.
At work we’re building a dashboard and it’s kind of a first for most of us. My colleagues let the copilot assistant do almost all the coding, then they have no idea how the system actually works. When time comes to do maintenance they can’t evaluate whatever may be going on and almost nothing’s planned as reusable components. It’s a loss for the company mid and long term
That means THEY (the CEOs) are still making shit decisions even though AI is providing quicker , more accurate, and higher quality information to them and their staff, or their company is so bankrupt on creativity and reliant upon their crappy inefficient corporate machine they made to realize the value and profit
AI is a long term investment. You have to be an idiot to expect profit short term. Especially when WE DON’T HAVE CAPABLE MODELS YET.
Ironically a bunch of companies forcing LLMs into their products, making their profitable companies become unprofitable and less userfriendly resulted in me using LLMs to just build self hosted replacement products for a bunch of subscription orientated web base apps, I imagine I am one of many, so not only are they losing money though their poor investment stratergies and ramming things people do not want down our throats, they are likely also losing money from competetor „legacy“ products that people are building to replace them,
44% seeing gains is pretty incredible! Lots of companies are attempting to roll out nonsensical AI features, etc. I would’ve expected only 10-20% would actually be doing it in smart enough ways to see benefits from it.
Check out for this project creation and deliverables based on user needs first three clients are free and there is 3 major categories for development of projects. Check out: [https://zyro-dev.vercel.app/](https://zyro-dev.vercel.app/)