While most have heard about Netflix’s successful AI engine, truth is that the majority of media businesses are struggling to accelerate their productive adoption of AI. Here are five key moves that they should consider when making those all-important use-case selections that will help ensure a successful AI initiative.
The convergence of data, technology and digital products is creating the fertile ground for AI initiatives to flourish. However, AI success is far from uniform within media companies. Most are facing one or more of a number of barriers and challenges, from having the right data platforms and architectures, through driving business value.
Here are five important steps to take to generate real value:
1. Identify cases that can deliver value quickly and that you can build on incrementally. ROI is important, but speed is critical. One common mistake that many organizations make in selecting the problem to solve with AI is to pick the most complex one – after all, the reasoning goes, AI is a tool for tackling complexities that the human mind can’t get a handle on. But that doesn’t necessarily yield the biggest or best payoff. There can be a far simpler problem which, if solved by bringing in AI-enabled automation or by leveraging an AI-enabled technique, can generate more value more rapidly. So it’s always wise to choose problems based on envisioned value rather than to focus on complexity. Return on investment must also play a role in decision-making. Under no circumstances should the cost of the solution exceed the value of the benefits, hence a fairly precise grasp of the expected ROI is extremely helpful. One way to help select the right projects is to factor in a higher cost of capital in your ROI models, as this will emphasize those that deliver value earlier while also reflecting the higher level failure risk of AI projects vs. standard technology ones. With AI, business validation is needed to decide whether a given insight will be at all helpful or not. Hence, it is common that some projects are scrapped after a technically “successful” pilot or MVP, if business leaders cannot leverage the outcome in the way they want.
2. Always involve the business. While this seems obvious, it bears emphasis: Staying connected with the business throughout the lifecycle of an initiative is essential. Trade-offs and decisions will be needed along the way. It is therefore critical to understand what the business needs to achieve, how important the problem is to them and, very importantly, how quickly they need it. Understanding all this and agreeing to the way forward will help to secure the necessary investment – and even increased funding if early results prove more promising than anticipated. Managing expectations is crucial too. AI can never guarantee 100% automation or 100% accuracy. So, it is very important to understand what a tolerable margin of error is for the business, and how the business plans to resolve errors or exceptions. This understanding clarifies where, how and to what extent AI can help the business and where leaders might need a fall-back plan.
3. Have the right data, at the right time, ready for analysis. Any savings that an organization can gain by using AI can easily be eroded or lost by the costs incurred to feed the data into an AI engine. As AI’s success is so heavily dependent on data, the data has to be the right quality, it must be trustworthy, and its inevitable biases must be well understood and mitigated so they don’t undermine the organization. The right data should be available at the time when it is needed (stale data can generate an insight, but that insight may be of no value). And it should be prepared, categorized and classified in a way that makes it readily analyzable. While media businesses in a digital world are flooded with abundant data, the inconvenient truth is that this data is often unorganized, unclassified, poor quality and/or out of date. So investment in modern data platforms and data management capabilities is critical to drive successful AI initiatives. An enterprise data strategy is needed that combines the data (i.e., via a data lake or the equivalent) and allows all areas of the business to access it freely. For example, a customer’s history on a directto-consumer platform is useful for product managers to understand what features are preferred, for customer service to know what the customer has done on the platform and for marketing to create personalized campaigns.
4. Start small, fail fast, be nimble. The phrase “first time right” does not necessarily apply to AI. That is especially true in making predictions and forecasts. Achieving an acceptable level of accuracy might take a number of iterations and continuous course corrections. Failure, then, must happen fast in order to learn what to correct. Since the stakes are high and there is always a risk of failure, it is also important to start with a smaller problem or a subsection of a large problem. This helps to reduce the risk associated with the cost of failure. There’s no shame in dropping an idea and rethinking the approach. In fact, that willingness to rethink is vital. If the viability of a solution is in doubt, persisting with it – and by doing so wasting time and money – is never the right way to go. It is always advisable to course correct or in some cases drop the idea altogether and pick up with a new one. AI is not a magic wand and it cannot solve every problem. Once a smaller problem is solved and the business can see its value – and associated ROI – the solution can be scaled up to solve a bigger problem.
5. AI projects are different from IT projects – so measure outcomes, not outputs. IT and AI projects are inherently different. IT projects proceed with a clear idea and a set target for the desired output from day one. AI, in contrast, is mostly used in the quest to understand the unknown. It’s therefore impossible to know what the output will be ahead of time. AI needs to be tuned, monitored and modified over time. Success with AI may require more iterations than originally planned, and it might not deliver the same level of accuracy or automation as initially imagined. Success should therefore be judged by the degree of impact it creates and how much value that yields. For example, a media organization spent eight months exploring multichannel advertisement impression data from 112 drama series to identify three impression pattern types. When advertisement impressions on 55 comedy series were explored, the same three impression pattern types came to light, proving that the segmentation criteria were able to scale and were generally applicable across different genres.
How to move forward then within your organization? Check out the recommendations in the whitepaper Five ways media companies can generate value from AI.