Next-generation AI-powered insights platforms are cutting-edge technology. As such, many businesses find themselves uncertain about how to choose the right one for their organization. Where are the current inefficiencies in finding information? What even are the pitfalls to watch out for? How do you even know if a platform is lacking a capability your organization would benefit from?
Fortunately, these four core questions can help clear things up considerably.
- Does this ai platform address your institution’s pressing pain points?
- How easy is this ai driver platform to deploy?
- What does the user adoption process look like?
- How will data scientists manage and kept their data current?
The following sections will highlight what a good answer to these questions should look like, as well as some concerns to watch out for.
1. Does the ai insights platform address your institution’s pressing pain points?
Identifying the pressing pain points that an insights platform can alleviate is a process. That process starts before deployment and should involve comprehensive help and guidance from the team behind the platform you’ve chosen.
Different businesses, and even different insights departments within the same business, will have different pain points. Here are just a couple of examples we’ve run into at Lucy:
- A Global Packaged Goods company found that the leading form of information gathering is over social channels like Teams and that 1000s of people rely on a very few SMEs to answer their questions. This made a huge impact on the SMEs – they spend so much time answering questions they can’t do the strategic work that was most valuable to the business.
- One office in North America conducted an extensive, multi-million dollar market research project. Only after the research project was nearly complete did they identify a nearly identical research project that was executed by their European branch the previous year. Ultimately, they discovered they were spending hundreds of thousands and sometimes even millions on redundant research,
Both of these examples required distinct functionalities to be solved effectively. Look for an insights platform with a team that is responsive, collaborative, and capable of creating a solution that’s fully optimized for your needs.
2. How easy is the platform to deploy?
Deployment shouldn’t be the hard part of adopting an insights platform. If it’s going to be an arduous process just to get the thing in place and operational, that’s ultimately low-value hours spent chasing a high-value dream.
Ask the team behind the platform roughly how long deployment could take. While, of course, an estimate is only an estimate, it can provide you with a general idea of what’s required. Some platforms need months or even years to be fully optimized, thanks to the complexity of their data ingestion processes.
That leads us to another consideration regarding ease of deployment: How much work will you and your staff have to undertake? If you’re looking for a solution that can provide accurate, comprehensive answers across your complete range of files and systems, what’s the process to get all of that into the platform? And can it interact with the documents and data you currently employ?
The data integrations process is often described as “garbage in, garbage out”. However, it doesn’t have to be that way. You can instead look for a platform that can index all documents, regardless of type or quality, and learn to spot what’s important. Such a platform eliminates the data grooming task and the human burden of maintaining the quality of your answers.
3. What does the user adoption process look like?
As anyone who has ever tried to implement a process change knows, user adoption is where the real work begins. Trying to shift ingrained behaviors in employees is a challenge. As bad as the status quo of searching is, it has been that way for so long that most simply aren’t aware of how much easier things can be. So your insights platform should be ready to help tackle this challenge head on.
First, the team behind the platform should be able to demonstrate to you that they have experience in successfully driving user adoption. It should be expected that your vendor will be actively engaged in the process, hosting trainings, providing resources, answering questions and troubleshooting problems if and when they come up.
Second, using the platform should be intuitive and rewarding. Look for features like:
- Questions can be asked in natural language. Instead of restricting users with the limitations of keywords, natural language processing allows them to ask their questions just as if they’re having a conversation with a colleague.
- The platform can be accessed from anywhere, anytime. Cloud-based insights platforms allow users across the world to ask questions and get their answers in real-time. This has become particularly essential over the past few years with the rapid increase in hybrid work.
- Answers instead of search results. Instead of receiving an endless list of documents that you are then compelled to painstakingly comb through, your insights platform should be able to bring you to exactly where the answer lives. And regardless of if it’s specific page in a document, a timestamp in a video or a line in a spreadsheet, the best answer should be ranked at the top.
- Works across all your systems. In order to truly get the best answer, your insights platform should be able to integrate with just about every database, every dashboard, and every filetype your company uses.
- The platform is easy to use. The more complex the user requirements are, the less inclined users tend to be to engage. Your insights platform should replace all the interruptions and low-level questions that would normally be asked of your subject matter experts. That means it must be as easy as it is to send an email or post a message.
When your insights platform has these features, most of the major barriers to user adoption are eliminated upfront. That means faster uptake. And the faster the uptake, the sooner that high-value functions can be prioritized.
4. How will the data be managed and kept current?
Finally, the data itself. As we’ve already mentioned, the data ingestion process can be incredibly costly if it’s not managed well.
Some insights platforms require you to curate and upload every document you want to include in the platform, making a file-by-file decision about what’s valuable for the platform and what’s not. That process continues even after the platform has been deployed. Decision-makers need to be trained in which future documents should be included and which should be ignored. It’s ultimately time-consuming, labor intensive, and often results in mistakes that can influence the quality of the results the platform returns.
To avoid these problems, look for a platform that, with machine learning, can automate all forms of data indexing that can read, watch, listen, and study every document to understand its value and rank it accordingly.
You’ll also be best served by a platform that can integrate into other systems, rather than requiring that everything be shifted onto it. For example, many organizations already have dozens or hundreds of SharePoint sites. Choosing a platform that goes to those sites and scans their documents in place saves time and ensures the platform stays up-to-date as new information is added.
How time becomes more valuable
When low-value tasks are streamlined and automated, insights professionals are freed up to focus on the high value. Tasks that solve problems, build relationships, develop products and improve profitability. Think about how many minutes per day each of your team spends looking for answers. Think about how often your subject matter experts get pulled off their strategic priorities to answer redundant questions, simply because people can’t find the answers. What does that add up to across your entire company, over months and years?
Insights platforms are a game-changer when it comes to eliminating low-value hours. But for your business to see the greatest reward, the platform you choose needs to address your specific needs without introducing new challenges.
The best insights platforms will be:
- Easily managed
- Easily accessible
- Intuitive for users
- Well supported by the business users behind the platform
- Easily integrated with your existing systems
And on the flip-side, they will not:
- Be burdensome to deploy
- Require labor-intensive data grooming
With an insights platform in place that meets these criteria, the value is tangible. Just add up all the time you’ll no longer be spending on low-value functions. That all becomes truly profitable time, time that can be used to make things happen and return results.