Why Data Science and AI are No Longer a Rich Brand’s Game
People have said it for years—data is the new oil, but refining it was costly. With AI tools making data science accessible, businesses of all sizes are tapping into smarter decisions and deeper customer insights. The big question is, do you build your capability in-house or buy something off-the-shelf? Tanya Hyams-Young, CEO of SourseAI, explores.
We’ve heard people say for ages that data is the new oil. They were right. It’s fairly abundant but realistically, very expensive to refine and turn into anything usable.
For many organisations, the whole data science wave had seemed like more trouble than it was worth. Yes, it would be good to refine processes or systems, analyse data, better understand customers and start to predict behaviours or results. But it required expertise, time and budgets that often weren’t available. Early adopters of data science in business typically invested heavily to reap the rewards, so it was mostly large organisations that were able to do it. Until now.
With AI rising rapidly, what was once available only to large organisations with vast budgets is now available to anyone. But it’s how you use it that separates the wheat from the chaff.
So now the decision to use AI and data science may be an easy one, the question of how to implement it is complex. Generally speaking, there’s two main routes a company could take: they could build their own data and AI capabilities in-house or buy off-the-shelf solutions.
Both approaches can deliver significant rewards but have traps for new players.
Let’s look at Woolworths. It invested $50 million in building an in-house AI and data science function. This has helped the company better understand and address everyday customer needs, implement dynamic pricing and a create highly successful customer rewards programme. AI and data science has become a strategic pillar of Woolworth’s business.
But big brands and budgets do not always mean success. Westpac learned this the hard way. It invested $70 million in developing in-house real-time decision-making capability, only to restart the project several years later due to various tech failures and challenges.
On the other side, innovative telco, MATE, was able to turn its abundance of data into specific business impacting campaigns, significantly increasing customer tenure and average revenue per user (ARPU), with an off-the-shelf solution. By partnering with an industry specific AI provider, it was able to see quicker results and embed its data function into all elements of the business, ultimately becoming and more dynamic and truly customer centric organisation.
There is much to consider when making the decision to build your AI and data science function or buy an off the shelf solution: desired outcomes or objectives, cost, customisability and much more. Building yourself allows you to ensure everything is specific to your organisation, its requirements and how it operates, but it’s expensive and time consuming. Buying a system reduces risks and allows for quick deployment and functionality, delivering faster value and accelerated time to market, but is less customisable.
It can be tricky to balance these factors, because deciding how to implement AI and data science requires specialised skills in data analysis, statistics, and domain knowledge. Unfortunately, many organisations undervalue this expertise and mistakenly hand the AI reigns over to the IT team because it’s ‘tech’. This almost always results in tools that aren’t quite fit for purpose.
Because AI alone isn’t a magic bullet, the first step is understanding what you want data and AI to do and how it will help your business achieve the desired results. Establishing a dedicated business case will help you recognise the strategic value of data science and help data and AI become a strategic business asset and not a side project. This exercise will help answer questions which will also help determine whether building an in-house function is more beneficial than purchasing an off-the-shelf solution.
The other important factor in the decision to build or buy an AI function is capacity. While AI has streamlined software development, many end-markets targeted by vertical SaaS do not have the technical talent or internal capabilities to build their own software. Industries like restaurants, construction companies, and not for profits are still early in their tech adoption phases and prefer off-the-shelf solutions tailored to their needs.
Lets look at MATE again. When it was looking to increase its data science capacity and use AI, it had a clear purpose in mind: it wanted to be smarter and more customer centric than all the other telcos. In practice this meant using data to better understand its customers, make better business decisions and deliver greater value for customers. It also wanted to move away from traditional telco sector acquisition practices, which it felt was leading to customer apathy across the category.
Tanya Hyams-Young
MATE also knew that any AI solution would need to integrate with some vital business tools – namely its custom-built CRM, its call handling software, Google Analytics and its marketing automation software. Ultimately, MATE went down the route of buying an off the shelf solution. It was able to partner with a leading AI decision intelligence platform designed and trained specifically for the telco sector. Now this system is a vital strategic tool for the business.
Co-CEO of MATE explains: “SourseAI tells the MATE team what it doesn’t know and uncovers specific customer cohorts that expect a different engagement experience to other customers. By creating personalised campaigns for these cohorts, MATE can communicate with the right customers, at the right time with the right offers to drive up their ARPU and increase their tenure. MATE is automating the messages to these cohorts through the integrated marketing automation tool that plugs straight into SourseAI out of the box.”
If he had one piece of advice from the process of implementing its system Mark Fazio: “it’s important to remember that AI is just a tool – not the complete solution. If you don’t have a clear purpose and plan for how you want to use AI, you’ll probably make bad decisions. Ultimately the AI is only as good as the information you give it. People are the strategy and key to success.”
While buying an off the shelf solution worked best for MATE, each organisation will have different requirements.
With the AI and data science landscape is accessible to organisations of all sizes, the most important thing is to do the strategic work upfront. Establishing a clear purpose, business case, and implementation plan that doesn’t overly rely on the IT team. With this groundwork, anyone can use data and AI to improve their business and stand out from the competition.
Keep up to date with the latest in media and marketing
