Spend Visibility as a Service: Searching for High Value Options


You should never be surprised how little technology many large firms have when it comes to procurement. The most foundational procurement technology is spend analysis. One of the problems in gaining spend visibility for companies that are early in their procurement transformation journey is getting the funding for this capability. It’s a real catch-22 situation. So how do you get this capability and not be doomed to building your own hobbyist tools using basic Microsoft desktop options?

Free Research: How to Wring Every Bit of Value Out of Spend Data

The dominant application in spend analytics by far is the Big A$$ SpreadSheet (BASS) system, an acronym I coined about 15 years ago that still lives on today. (Even Ramsay Chu, Rio Tinto’s CPO, has used it! Feel free to use it anytime – always gets a good chuckle with an audience.)

So say you have about $20,000 to $40,000 available to stand up a basic capability in spend analysis. What should you do?

Look to the cloud, of course!

But what are the options? You can get some querying and reporting with a slick tool like Qlik or Tableau. (You can see some Qlik examples here, here, here, and here.) This is great if your IT department uses one of these tools, is looking for something to do and won’t charge your cost center a lot. But even if you’re the 5% of companies to whom this might apply, there’s a bigger problem – the crappy data. Sizzle is great, but sizzling garbage is still just garbage. The cloud is also great, but empty apps that don’t deal with the data problem are just going to slice and dice bad data.

So, how do you solve this problem?

First, realize that basic spend analytics generally has 3 components: the operational data store (ODS) to collect the data from all your source systems; the transformation layer, which is the cleansing, normalization and classification process; and the reporting layer, which also includes ad hoc queries, discovery and dashboards. I talked about the third layer above. In terms of the data store, I personally advocate building your own – although third parties can help – so that you can have a single source of data to integrate to the next 2 layers.

The real magic in spend analysis though is the transformation layer. Yes, you can do some transformation as you load your ODS and maybe even, God forbid, fix some of the data at the source. The magic isn’t just the empty analytical software -as-a-service (SaaS) application, but rather the data transformation performed as a service that feeds the cloud application with good data.

The market needs a robust, accurate and value-priced auto-classification-as-a-service capability. I’ve written about this general problem before, but given the price point I mentioned above, it can be a challenge. Even strong players such as Spend Radar, Rosslyn Analytics and Opera Solutions, which owns BIQ, will only scale down so far price wise. So, if you want such a capability, you’ll need to get creative, think through how your solution will grow over time and decide how it fits into your broader procurement transformation.

For example, if you need professional services support to augment the spend analysis with category analysis and intelligence, consider a professional services firm that embeds a commercial solution as part of its diagnostic offering. Or pick a services firm that offers a native spend visibility tool. Insight Sourcing Group’s SpendHQ solution is a good example.

If you want a provider that has offshore support to help you with your data wrangling – which is no longer a problem, as long as the labor is used to train and feed a software tool to make the process more repeatable over time – then some choices exist there, too. For example, Enrich is a provider that is a no-brainer to consider for such as a classification-as-a-service capability in a company that is an Oracle shop.

There are many more options, and I encourage providers to chime in below in the comments section. But if you have any questions and want to discuss this topic, whether you’re a provider or a practitioner, please don’t hesitate to reach out directly.

Voices (3)

  1. Ronald Kok:

    Hi Nadine,

    I am working with SAS Netherlands to develop a custom spend analytics solution based on SAS VA, for about 10.000 invoices per year.
    I very much like to understand what your experiences are what kind of problems / challenges you encountered.
    Can we have a chat?

  2. Pierre Mitchell:

    Thanks Nadine, yes, let us know, especially about what types of spend category classification you can do natively within it. SAS has a spend analysis solution, but it’s been a non-player for almost a decade, and it’s too bad (SAS needs a new leader in that are like Denny Norman was) because it had some nice capabilities. I’d love to see it moved into the cloud with a Visual Analytics cloud-based front end.

  3. Nadine Sauve:

    I’m working with a client who has 2M rows of AP data per year. Excel and Access can’t handle it. They freeze and crash running simple classification rules. Client has limited budget <$20K, no IT support and less then 10 analyst who need access to the spend data for reporting. We're testing SAS Visual Analytics, cloud version. We haven't launched yet but preliminary "trial" was positive. Interesting component is the forecasting abilities and advanced analytics potential. If we can improve our input data, we may have an opportunity to take our basic spend analysis activities much further and bring a whole new scope of value to our strategic sourcing and supply chain.

Discuss this:

Your email address will not be published. Required fields are marked *