Our first Question of the Week – driving procurement benefit from Spend Analysis

We explained earlier today that we're starting a regular new feature - the (procurement) question of the week (QOTW) - and here is the first.

We received this question the other day as a comment on a very old post. It's from Andrew who works for a UK local authority and commented that his organisation was using one of the major procurement solutions vendors that we had featured ina  a post (he didn't say which one) to deliver spend analysis. This is his question.

"I would be very interested to capture views as the best methods/approaches to utiilising the spend data that is eventually captured. Can anyone give me their views as to the best approach they have used in identifying savings opportunities through better spend analysis?

A very good question; I know from experience it can be quite difficult when you're first faced with a load of data emanating from this process, particularly if you haven't had good spend analytics previously.

So we'd love to hear both from practitioners who have been through this already; or providers of Spend Analytics solutions in terms of what they would suggest as the early benefits that a user should be looking to develop. We'll feature responses - and our thoughts - on Friday.


Voices (12)

  1. Simon Dadswell, PROACTIS:

    Spend Analysis – Don’t Get Blinded by Data!

    Enterprises are often confronted with significant data availability and quality issues when it comes to achieving visibility of spend to identify off-contract purchasing, track and enforce superior spending practices and address process inefficiencies.

    For many, the response is to hire expensive outside consultancies to address these problems. This results in large-scale data cleansing and enrichment exercises, taking months to deliver a snapshot of analyses that quickly become obsolete.

    Issues of poor spend information include:

    – Disparate data: Spend data is scattered across multiple, disconnected systems (ERP, AP, eProcurement systems, purchasing cards etc.), territories and sites.

    – Inaccurate data: Spend data is often recorded inconsistently with errors and duplicates. Suppliers are often spelled differently and there is no link between their parent company and subsidiaries.

    – Incorrect naming: The use of terminology varies between business units and there is an inconsistency in naming practices for suppliers, products and services.

    – Limited analytics: There is not enough meaningful information held in the General Ledger and cost centres to drive procurement decisions etc.

    While, the objective of creating good, well-structured information is admirable, most procurement teams need only take a “back-to-basics” spend analysis (or procurement review) to uncover huge reserves in potential savings.

    They do not need to get lost in or be distracted by data as the only means for making sound decisions. Nor, should they allow the lust for ever-more granularity of spend data to overwhelm all sensibilities. Data itself doesn’t make decisions or come up with good ideas: people do.

    Start with a rapid two/ three-week exercise to establish the critical information about the enterprise and its spending behaviours. Review the ‘as is’ position within the spend control process, and use external experts with the right tools to help identify relevant process and policy changes, and derive a sense of magnitude with regard to improvement opportunities.

    Don’t dig too deep or turn the analysis into a never-delivering consultancy exercise; keep it short and sweet. Limit the scope if necessary. The review should include a discussion and review of existing data (noting its limitations to be addressed at a later date), uncovering key facts and detail surrounding:

    – Process (supplier management, sourcing, ordering, invoicing and on-going contract management)

    – Supply base, spend details and transaction volumes

    Many of the answers quickly become apparent. A rapid exercise of this nature can easily highlight real benefit potential, in terms of efficiency, cashable benefits and risk mitigation, and prioritised with details, scheduling and resource implications.

    Consider: according to analysts, less than 20% of enterprises are utilising the full disciplines of spend control – let alone exploiting eProcurement tools to better automate these processes. This equates to substantial opportunities for savings (both cash and efficiency).

    By pulling back the cover on spend, enterprises can find ways to reduce spending by up to 15% and cut cycle times by as much as 80%.

    – Improve negotiation leverage (use aggregated demand to negotiate better prices with suppliers and rationalise differences in prices charged by suppliers)

    – Consolidate suppliers (consolidate the supply base to select preferred suppliers and leverage the increased spending with suppliers)

    – Bring more spend under management (conduct sourcing events for the category to establish approved pricing and suppliers to realise savings)

    – Reduce maverick spend (reduce off-contract spending, realise savings on spend shifted to approved vendors, and reduce purchase-to-pay cycle times)

    – Improve budget compliance (manage comparisons between budgeted, actual and committed spend and ensure purchasing compliance with organisational polices)

    – Increase invoice matching performance (achieve straight-through processing of invoices to improve productivity and reduce costs) etc.

    PROACTIS, a leading spend control and eProcurement vendor, is an advocate of such an approach and argues that “spend control is a journey of iterative process improvements supported by automation”.

    Recently, PROACTIS was commissioned by a UK public sector organisation to undertake this style of exercise. In just a few weeks, they identified in excess of £566,000 savings per annum (in isolation to process efficiencies) and put in place a set of action-orientated recommendations.

  2. Luke Spikes:

    A great question and some very through provoking reponses. I particularly concur with the comments that relate to ‘doing something with the data/results’ – it is not uncommon for us to deliver ‘spend & contract visibility’ in the form of the data and online tools to be asked by the head of procurement ‘so what do we do now then’ (NB – we work exclusively in the public sector). To provide a ‘way pointer’ for the procurement manager seeking inspiration we’ve developed what we refer to as the 10 Point Plan refer http://www.spikescavell.co.uk/products/analysis/10pointplan.aspx for a description). The 10 Point Plan has been derived from direct observation of the results of our public sector customer’s efforts to transform the way they procure goods services over the past 6 years or so and is, in essence, a set of procurement improvement initiatives which, if successfully executed by the procurement function, have been proven to have the potential to deliver savings (provided of course you have the spend analysis to hand !). To the more experienced practitioner it will doubtless smack of grandmothers & eggs but in our workshops and seminars when the 10 Point Plan is introduced it ALWAYS results in significant uplift in surreptitious note taking, no matter how expericienced or senior the audience !

  3. Graham Seddon:

    Spend Analysis will produce the data required to form a strategy for savings, although the only system available which will provide Procurement control and visibility is Fusion by OCG. Anyone whom would like more information please e-mail me for details.

  4. Mike Roberts:

    I agree with the points above about ensuring that the analysis is acted upon, and the Kraljic matrix concept that puts additional layers of interpretation over the actual spend data to assist in prioritisation etc.

    But if I may offer a practitioners view on getting value out of your spend analysis… I believe the key point is to match the time and effort of gathering the data and sorting through it, with the level of detail/quality of the data and the type of analysis required in the organisation.

    For example… if no clear and actionable spend analysis has been previously performed in an organisation then the ‘quill pen and ink’ approach might offer the best return on your efforts. A series of quick win opportunities such as overly fragmented categories may be quickly identified, and provide a sourcing team with several months of work.

    However, if this simplistic situation is complicated by the sheer volume of data/number of systems in which the data is held, the detailed level of classification required in certain categories, or the fact that the quick win opportunities have all been found, then a more robust spend analysis process might be required.

    The process must be able to rapidly refresh data to keep the analysis up to date, report on various metrics as from as broad spend by supplier, to complex ones such as standard vs actual unit price by material. It also must be able to handle the timely update of data anomalies (they will always be there!) such as classification updates or supplier groupings.

    Once the process is running well then additional views of the data such as contract compliance, supplier performance and savings reporting can be added over the top. When tracked over time these metrics really provide a view of whether the implemented procurement strategies are delivering.

    So I agree with starting with quill and ink, but be ready to move to more sophisticated methods as the business demands!

  5. Matt McGovern:

    “If you don’t measure what you are doing, how well you do it, and whether it’s getting you to the place you want to be, how do you know if your efforts will yield success?” That’s the position of Harvard Business School Professor Tom Davenport, the author of the book Competing on Analytics (ISBN-10: 1422103323). In that book Davenport address this issue which is the same for all users of data analytics not just Spend Data.
    Davenport identifies what to look for in data analytics. It should be Data rich and information intensive so you can go deep into the data to see hidden patterns. Asset intensive – meaning you can monitor feeds from lots of systems to see the whole picture. Web enabled, allowing distributed decision making from a common view/same information. Finally the data should also be Dependent on/linked with the process you are monitoring so you can see exactly what’s not working.
    So how do you use that data once you have it, how do you use it to your advantage? Davenport suggests the data can be used for areas like; Modeling allows us to look at past data to tell us what we learned. For example, did we buy less consulting services? We can also use the data for Alerting and recommendations, to help define what the next best action is to solve the problem.
    Of course the key to using this data in procurement is for Optimization and prediction analysis thus providing insights about the future – what are the best and worst outcomes – using “what if” scenario analysis. For example, how much will I pay not to have to put all my eggs in one basket by sole sourcing a commodity?
    love to chat more on this with anyone interested…

  6. Kirit Pandit:

    In our experience, many of our F-500 customers have identified different sets of opportunities based on where they are in their procurement transformation journey.
    1. Phase 1 – Visibility. This is for customers beginning the transformation journey. Typically, these organizations do not even have good visibility into their spend. So the primary objective in Phase 1 is to provide them 360 visibility so they can identify the 2 most basic types of opportunities – Vendor rationalization and Demand aggregation. These are the low hanging opportunities.
    2. Phase 2 – Maverick Spend. This is spend leaking out because of violations of the procurement process. POs being created without a contract. Invoices cut without POs. Payments made without invoices, etc. Also, mismatch of payment terms is a big one.
    3. Phase 3 – Contract Compliance. This is the Holy Grail, if the customer is past phase 1 and 2. This is also the most elusive opportunity for a variety of reasons.

    We also utilize the flexibility of our solution to bring in other value-added information and marry it to the spend data so the users can conduct a more thorough and impactful analysis. This information includes items such as the “Date a category was last sourced”, “The Criticality or Risk level associated with a Category”, “Single or Sole Source Category”, “Expected Impacted of Savings program”, “Complexity or organizations ability to conduct a saving initiative in the category”, “Contract Expirary Dates”, “Supplier Risk/Credit scores” etc. Much of this information can be obtained by examining historical events and conducting buyer interviews. Combined with spend information, it can help you figure out which categories or contracts or suppliers to go after.

    Kirit Pandit
    Product Management, Spend Analysis, Emptoris

  7. Flog:

    I agree with Christine and use the Kraljic matrix however, and I don’t want to sound negative, I find that some people find the concept of using numbers and 4 boxes far more complicated than it actually is. I was using spend analysis to very good effect back in the early 1990’s by simply identifying my top, say, 200 or so suppliers, classifying them into commodity groups and then looking at (a) how much we were spending on a commodity (b) the number of suppliers and (c) using my personal knowledge of what was happening in the organisation, why we were using them. A simply calculation showed me which categories I had under ‘reasonable’ control and where I needed to concentrate my efforts in the short-medium term. It was basic and it worked.
    Today, there are excellent spend analysis providers which provide a wealth of information, the ‘quill pen and ink’ analysis I did over 20 years ago (am I that old???) will still work; however, the availability of the additional information, size of organisation, turnover, geographical location etc can provide you with real background information upon which to develop sound procurement strategies. You can see the likely impact of going for one mega-contract compared to, perhaps, setting up a number of separate ones with the chance of smaller SMEs having a chance of winning. At the end of the day they may not win the contract, but at least they’ve a chance of getting into the game. I’m not talking about disaggregation to ‘avoid the EU rules’ but creating sensible contracting opportunities that will present real value for money solutions.
    To anyone getting into spend analysis, remember it’s a snap-shot in time, use the basic data, be selective in what you look at; then, once you’ve decided on a high level strategy, interrogate your financial system to update your understanding and data if only to make sure that something key hasn’t changed between the date of the spend analysis data and the present. Then move forward.
    However, don’t spend hours ‘playing’ with the data and doing nothing with it.
    A number of years ago I was in a meeting with a Finance Director (from another organisation but within the same sector), he received his organisation’s analysed data and the conversation went something like –
    He: ‘..very interesting, the things I learnt about what was going on’
    Me: ‘So, what did you do with the information?’
    He (looking confused): ‘What do you mean?’
    Me: ‘Did you change anything? Set up any new contracts, investigate any of the expenditure?’
    He (looking confused): ‘No’
    So, do use spend analysis – keep it simple at the start and ‘grow into it’. Use a service provider to do the number crunching, but, however you do the analysis – do something with it promptly (not 11 months later)

  8. hassan iftikhar:

    Spend analysis realy helps.But what constitute to be a spend data.What is the intergrity of the data.how it is calculated.Is it in the right format which can provide us meaningful inferences.Unfortunately this fact takes a lead position before we talk of spend analysis and its benefits.Especially in SME’s this credibility of data and the way it is captured by Accounts Function and its further integration with procurement needs of spend data is a big ? I would like to see some responses on this issue as well.

  9. Pete Loughlin:

    Great question.

    Spend analysis is a fundamental procurement tool without which it is impossible to prioritise; impossible to manage risk and impossible to know if you are regulatory compliant but I’ve seen spend analysis used almost as an end in itself. Great analysis, great charts and diagrams that describe the state of procurement – but so what?

    Looking forward to seeing the ideas later in the week.

  10. Christine Morton:

    Kraljic matrix to the rescue!

    Map the spend ino the Kraljic quadrants on the 2 x 2 matrix – one axis being value, the other risk, or criticality to service delivery. Identify which categories of spend fall into the strategic, leverage, bottleneck, and routine. Then identify which procurement savings strategy is most appropriate based on its quadrant.

    The choices for savings strategies include joining a collaborative contract for the routine stuff like stationery, all the way up to supplier relationship development for the strategic services like waste collection contracts and PFIs.

    Good spend analysis will help you identify the categories involved and which of your suppliers fall into those categories. It will also allow you to drill down into the detail to find out where funds may have been misspent (oh we all remember the stationery exercises from early in our procurement careers and having to tell senior executives that no, they don’t need the fancy schmantzy weighted paper).

    However, what spend analysis from such companies is unable to do, in my opinion, is assign a value as to the criticality for your organisation. They can tell you how much has been spent, but they can’t tell you whether it’s strategic to service delivery. Use of the Kraljic matrix, however, will ensure that you don’t spend too much time worrying over the office furniture contract when there’s bigger issues in the highways contracts (just as an example). Apply resources as required.

Discuss this:

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