--- title: The success of your AI program is decided long before you choose the technology description: What data assets can telcos monetise now to unlock new revenue streams? SourseAI CEO Tanya Hyams-Young explores this timely topic. --- [SourseAI Blog](https://blog.sourse.ai) # [The success of your AI program is decided long before you choose the technology](https://blog.sourse.ai/the-success-of-your-ai-program-is-decided-long-before-you-choose-the-technology) Written by [Tanya Hyams-Young, CEO](https://blog.sourse.ai/author/tanya-hyams-young) | 10/11/26, 12:20 am In this article I want to share what we see at Sourse when AI delivers real value, and what leaders can do in the next quarter to tip the odds in their favour. We will cover four things: - Why culture and leadership matter more than the model - How governance helps move teams faster - How to measure the value that you set out to create - Practical places to start ## Introduction When an AI initiative stalls, the first instinct is to look at the technology. Was the model good enough? Was the data clean enough? Did we choose the right platform? In my experience those questions matter, but they are rarely the reason things go wrong. The more common cause is that the organisation was not ready to change how it works, and no one was clearly responsible for making that change happen. ## Why culture and leadership matter more than the model Most organisations can now access capable AI tools. That means the technology is no longer the difference between the companies that benefit and those that do not. The difference is whether people trust the output, whether they have permission to change how they work because of it, and whether their leaders visibly use it themselves. We see this clearly in marketing mix modelling for telcos and MVNOs. The modelling itself can be run in hours rather than months, and it can help a marketing team stretch its budget 25 to 40% further. But the analysis only creates value when the marketing team is willing to move budget away from channels they have always relied on, and when the CFO and CMO agree on how they will judge the result. The statistics are the faster part. The conversation about what to do with them is where the value is won or lost. Leaders set the conditions for that conversation. When a leader asks to understand the data behind a decision, welcomes a finding that contradicts the plan, and rewards a team for stopping something that is not working, people learn that AI insight is safe to act on. When leaders say they support AI but keep making decisions on instinct, people quietly do the same.   ## Governance is what allows teams to move faster Governance has a reputation for slowing things down, and badly designed governance does exactly that. Good governance does the opposite. It answers the questions that otherwise stop every project at the same hurdle: who owns this data, who can use it for what, who is accountable if the output is wrong, and when does a human need to review the decision. There are two layers worth keeping separate. Data governance is about whether the information feeding your AI is accurate, well described and used appropriately. AI governance is about how the outputs are used, monitored and challenged. A simple, clearly owned version of each is far more useful than a long policy that nobody has read. A good test is whether a team with a new idea can find out within a day what they are allowed to do and who to ask. If the answer is no, the lack of clarity is already costing you momentum.   ## Measure the value you set out to create Many AI projects are measured by activity. How many models were built, how many people attended training, how many pilots are running. These numbers feel reassuring, but they do not tell you whether the business is better off. Value realisation starts before the project does. Agree on the commercial outcome you are aiming for, such as lower churn, a better customer lifetime value or reduced handling time in the contact centre. Agree on how you will measure it, who owns the number, and what the baseline is today. Then check it regularly, in the same forums where you review the rest of your performance. In the subscription world, we encourage leaders to anchor on customer lifetime value, because it forces the whole team to look at the quality of the customer base rather than simply its size. The same principle applies elsewhere. Choose the measure that reflects value created, and make it visible.       ## A practical starting point to move fast You do not need a grand transformation program to begin. These steps are achievable intra quarter: 1. Choose one business outcome that matters and name a single executive who owns it. 2. Pick one or two use cases directly linked to that outcome, and be honest about the data they need. 3. Write a brief statement of who owns the data, who approves its use and when a human reviews AI decisions. 4. Set a baseline and a target before any build begins, and agree how often you will review progress. 5. Run a short session with the relevant team (ideally the leaders in that team) on how they will use the results in their own decisions. 6. Share what worked and what did not with the wider team, so that learning spreads and confidence builds. The most important thing required here is time from leaders, and that is the point.     ## The technology will keep changing, and the fundamentals will not Tools will continue to improve quickly, and that is a good thing. What will not change is that people adopt what they trust, leaders shape what is normal, and value is created when insight turns into a different decision. Organisations that invest in those fundamentals will be ready for whatever the technology does next. If you would like to talk about how this applies in your business, or how we have approached it with telco and other clients, I would be delighted to hear from you.     [View full post](https://blog.sourse.ai/the-success-of-your-ai-program-is-decided-long-before-you-choose-the-technology) ```json { "@context" : "http://schema.org", "@type" : "BlogPosting", "author" : { "@type" : "Person", "name" : "Tanya Hyams-Young, CEO" }, "dateModified" : "2026-11-10T00:20:19.005Z", "datePublished" : "2026-11-10T00:20:18Z", "headline" : "The success of your AI program is decided long before you choose the technology", "image" : { "@type" : "ImageObject", "height" : 1024, "url" : "https://27114550.fs1.hubspotusercontent-eu1.net/hubfs/27114550/AI-Generated%20Media/Images/Create%20a%20landscape%20editorial%20hero%20illustration%20for%20a%20blog%20titled%20The%20success%20of%20your%20AI%20program%20is%20decided%20long%20before%20you%20choose%20the%20technology%20Visua.png", "width" : 1536 }, "mainEntityOfPage" : "https://blog.sourse.ai/the-success-of-your-ai-program-is-decided-long-before-you-choose-the-technology", "publisher" : { "@type" : "Organization", "logo" : { "@type" : "ImageObject", "height" : 60, "url" : "/hs/hsstatic/content_shared_assets/static-1.4092/img/default-amp-logo.png", "width" : 60 }, "name" : "SourseAI Blog" } } ```