February 3, 2025

Finding The Right Problem To Solve - Hang On... I Know There Is A Document On This From Last Year

Why Companies Often Solve The Wrong Problems

Identifying which problems are worth investing in is not as easy as it seems. Often it feels easier to prioritise something because a supplier is offering a snazzy solution. But what if you can’t shake the feeling that this doesn’t solve the problems that really plague your teams? What is the informed alternative?

Improving information-retrieval is exactly the type of problem that few companies tackle head on. Spending executive time thinking about the tedium of searching for documents is often the route to problem-solving enthusiasm. Yet it is precisely in eliminating this everyday drudgery where real value often lies.

Is This A Big Enough Problem To Solve?

How do you know if solving this problem is worth the effort? What if the uninspiring nature of the problem is because the real value lies elsewhere? Like other projects, the answer lies in applying the same assessment of Return on Investment (ROI that you’d use elsewhere). Focusing on concrete metrics like this, rather than unquantifiable promises like 'greater customer satisfaction', ensures you can truly compare the costs and expected savings. 


Before diving into spreadsheets, start with something more intuitive. Look at the gap between your most knowledgeable worker and your newest employee. Compare how quickly your team leader finds information versus a new hire. Part of the difference in performance (and possibly a large part) is attributable to knowing how things are done and often who to speak to. The variation could be down to issues in your workflows and tools (system problems) as much as differences in skill of the employee. If this gap feels large, this could be an indication that there are gains to be had from investing in employee familiarisation, or by improving the tools they use to navigate the organisation. 

A Real-World Example: When Poor Information Retrieval Costs You

Let's look at a case where information retrieval meets the threshold to try and fix it - at least to work out how much a solution would cost! 

A common challenge among our clients is that executives and sales teams struggle to recall previous projects in time to support new bids and pitches. Whilst they know dozens of previous client engagements could serve as perfect case studies, they are left relying on memory alone. 

Investment firms face a similar challenge: they sit on decades of memos and internal documents rich with cross-industry insights justifying previous decisions, yet accessing this knowledge remains difficult. 

Ironically with more experience the accessibility of any particular insight becomes harder to retrieve.


Beyond Keyword Matching: A Modern Approach

The traditional solution involved layering on additional search-accuracy features, cross-software integrations, and comprehensive knowledge management tools. Nevertheless, in most organisations, retrieval bottlenecks persist. The modern solution lies in semantic search - a method that understands the intent and meaning behind a query, rather than just matching keywords. Unlike approaches - which still exist in many organisations - that demand exact word matches, semantic search understands context, synonyms, and relationships between concepts. It’s similar to how the AI in tools like ChatGPT can understand what you mean, and rephrase it in totally different words. Semantic search addresses two key flaws in the searching status quo: rigid software-information silos and the reliance on exact phrasing. Even better, it also plays to human strengths, our memory. 

Even vague recollections or partial details can lead you directly to the sought-after information. A great illustration of semantic search in action can be found in this cheap and quick project we came across on Twitter, now open-sourced so anyone can use it. The image shows how semantic search can seamlessly navigate over 5 million text-data points (Tweets) from 250 sources to find what the author was looking for. These same capabilities extend to PDFs, images, and spreadsheet content - as presumably your organisation doesn’t run on tweets! 

For our clients sifting through years of documents, at decidedly human speed semantic search offers a potentially simple route to immediate gains. Gains which will be felt by frontline teams, where adoption and buy in are key. If this sounds like you, it’s worth assessing whether improving document retrieval would meet your ROI threshold. 

Choosing the Right-Sized Solution for Your Organisation

You might not need to implement much (or any) new technology at all. Sometimes the most practical solution involves spotting common queries and putting the answers in easy to find places, like an FAQ document or presentation of greatest project hits. These quick fixes can deliver significant time savings across your team. Sharing bookmark collections, documenting search tips, creating simple redirect guides: often, these collaborative solutions can address the problem without major investment.

For larger organizations or more systemic issues, comprehensive solutions like semantic search may be justified. Let ROI be your compass in choosing the right scale of solution to ensure the true result is sustainable financial impact. 

Technological change needn’t always - in fact rarely should - involve a new complicated system that looks like magic. Often the best problems to solve are the ones teams face everyday. The first step is to watch how teams are actually spending their time. Speeding up steps of this - particularly the ones that feel routine and uninteresting - is often the best target, attention grabbing or not. 

Related posts

AI Agents and navigating the web

The underpinning technologies that drive the internet, or rather that drive the profits of the biggest companies on the internet, go a long way in determining what it feels normal to do online.

AI Investment and Sovereignty: Rethinking What Counts

The future of British sovereignty in AI will not be decided by where our servers sit. It will be decided by who builds, controls, and profits from them. The investment decisions that determine this are being made now.

How an AI workshop can unlock adoption across your office

How to run your own AI workshop. Pre-workshop: don't assume anything

Beyond the Billable Hour: How AI and Client Demands Are Reshaping Professional Services

For decades, legal, consulting and accounting firms billed by the hour. Labour time was linked directly to revenue. The approach worked when time was the main input and when work moved at human speed. In 2025, AI is beginning to change and weaken the link between worker-hours and output.

No items found.