Turn Procurement Analytics into Real Business Impact

Many procurement teams feel like they have data, but can’t trust it. They’ve invested in platforms and performed analytics, but the resulting savings don’t meet expectations. CPOs aren’t short of reports; they’re short of confidence in the underlying data. They need analytics that surface insights based on relevant data they can trust, contextualized into recommendations they can actually act on.

WNS Procurement’s Vice President, Sourabh Gogna, and Procurement Analytics Leader, Disha Kumar, joined James Moore, Editor of CIPS Download, in a recent webinar. They discussed how combining AI and human intelligence turns procurement analytics into real business impact.

WNS’s framework for turning procurement analytics into business impact follows a four-step process — analytics, augmentation, action, and advantage. It’s crucial to move through each layer to successfully translate analytics into action that delivers real outcomes.

1. Analytics

   Your data foundation, pulling information from all your systems to understand what you have, and what needs enriching.

2. Augmentation

   The human plus technology layer, where experts turn analytical output into something that’s genuinely usable.

3. Action

   The decision itself, made as a result of trusted data outputs and supported by human context and judgement.

4. Advantage

   The compounding benefit you achieve over time as the cycle keeps spinning.

Watch the on-demand webinar to see exactly how this process plays out in real-world scenarios. Or read on for the highlights.

In many procurement teams there’s a gap between data and decisions, and it comes from a breakdown in communication. Often, analytics teams and category teams aren’t speaking the same language. Their capabilities have developed in parallel, but not in harmony.

This creates an analytics team that’s technically strong, and produces solid outputs, but works a distance away from where sourcing decisions actually happen. And it fosters a category team that’s commercially sharp and used to making decisions based on their own experience. They don’t fully trust analytics outputs because the models don’t reflect what they know about the supplier, the market, or their relationship.

By bringing category insight and market data into analytics, teams can produce the relevant, timely outputs that category managers need to make confident decisions.

But what does this look like in practice?

Strong analytics capabilities based on accurate, relevant data are fundamental to surfacing recommendations procurement teams can action with confidence.

For example, before you can benchmark contracts or model costs, you need a reliable and structured view of what you’re spending, who you’re spending with, and what you’re spending on. This will be the foundation from which you can build your advantage. But this is also where many organizations underinvest.

WNS worked with a large FMCG company with roughly £800M in annual procurement spend and operations across multiple sites. It buys seasonal commodities such as grains and sweeteners where timing and volume decisions directly affect cost. This company’s issue wasn’t a lack of data, but fragmentation across its systems and latency in getting usable outputs. By the time data was reconciled, category managers were reacting too late.

We built a spend-analytics and budget-control capability, consolidating and cleaning three years of historic spend data and connecting it with other internal sources to create a demand-forecast table. We also added the company’s contract database and linked everything to near real-time market-price intelligence.

Then we built a decision tool to give category managers recommendations for how much to buy, from which supplier, and when. All based on inventory levels and market conditions. This created proactive spend management within the company. Category managers stopped reacting to market movements and began anticipating them with a forward-looking intelligence system built on properly classified and connected data.

The most common gap organizations face in the analytics, augmentation, action, and advantage process is in the augmentation layer. They jump straight from analysis to action without giving an expert the opportunity to interpret, contextualize, and strengthen the output. And without this step, the whole process can break.

Augmentation puts a procurement expert between an AI model output and a decision-maker. They review what the AI model surfaces, such as cost anomalies, contract gaps, or savings opportunities, and ask the questions the model can’t answer. Is this real or just a data artifact? Is this the right moment to act? What external factors should inform my response? Who are the right stakeholders to discuss this with?

Procurement experts also add contextual intelligence AI models don’t contain including market benchmarks, supplier financial health, geopolitical risk, and commodity indices. They connect internal signals to external context and turn it into defensible and actionable recommendations. This bringing together of AI outputs with deep human intelligence and judgement is the essence of augmentation.

For example, WNS built and ran an invoice-control tower for a global food manufacturer, processing thousands of invoices each month across multiple sites and over 200 suppliers. AI automated about 90% of the invoice ingestion, looking through formats and languages, comparing each line item against contract rate cards, and classifying invoices as matched or mismatched.

The 10% of invoices that hit the exception queue — and the most commercially significant invoices in the portfolio — were investigated by a WNS procurement specialist. They looked at why the mismatch occurred, what it revealed about that supplier’s billing behavior, and how that compared to similar suppliers in the same market.

As well as using these findings to resolve compliance issues, they informed quarterly commercial recommendations. This is what augmentation should look like in practice. AI is there to handle data at scale, and human experts turn the exceptions into strategic commercial action.

The organizations that get the best actions and, therefore, advantage from their data, are the ones that treat analytics as a decision-support function rather than just a reporting capability.

A clear example of an organization that does exactly that is Merck, a long-standing client of WNS. We previously partnered with this large pharmaceutical company on a single, critical problem: tail-spend management. Like many large organizations, Merk had grown through acquisitions and amalgamation and ended up with multiple ERP systems, S2C platforms, and CLM tools, with data scattered across silos.

It needed external support to unify this data and execute decisions at scale. This wasn’t because Merck’s internal team lacked capability, the leadership team had made a deliberate choice to keep internal talent focused on strategic priorities. Our specialized team could interpret data and take action to handle the large volume of transactions under €100,000.

Over a six‑year partnership, we managed about $1 billion in spend, delivered 7–8% annual savings (around $50–60 million in yearly cash‑flow impact), and reduced turnaround times to under two days. This turn-around time is exceptional given the volume of transactions involved. The secret? Treating data analytics as a central function, augmenting data with external intelligence and empowering people who work with data to take action on it.

This is just a brief overview of the topics discussed in the webinar. To explore them further — and get insight into how the analytics, augmentation, action, and advantage process can help you in other areas like scope 3 emissions — watch the on-demand webinar.

Or if you’d like to discuss how to convert your own analytics into tangible procurement outcomes, get in touch.