After few years in SAM, I realized one thing – it’s hard to explain why SAM is so difficult, why it requires so much effort and why it’s taking so long to get perfect results. Both my friends and colleagues who are not involved in SAM keep asking questions like: Is SAM really a full time job in your company? What is so difficult in comparing purchased licenses with installed software?
Sometimes I try to explain and sometimes not, but such feedback always triggers my own thinking and I keep asking myself: must it be so complicated to implement and maintain SAM in a large, international company? Unfortunately, I haven’t yet found a magical way to make it really quick and simple; but I think there are some ways to make the journey a bit faster and more successful. You can look at SAM implementation from multiple perspectives and I’d like to choose two that I see as critical for success: governance and data. These aspects are important in general but even more important for companies with non-centralized operations.
I’ve learned that correct governance is extremely important in large and international companies. In my company, which has almost half million people across the globe, there is no chance that things will go right spontaneously. You can have perfect technical solutions, strong policies and excellent ideas, but implementing those consistently across all places and organizations is impossible without a strong governance model.
Governance should include certain aspects that I recognize as highly important for SAM within a large company:
SAM doesn’t exist in isolation. There are typically many different organizations involved in (or impacting) SAM in a company and most of these organizations wouldn’t see SAM as their main business and priority. For example, license balances will reflect changes that are performed by many business projects or ongoing IT infrastructure management and optimization. It is common for some actions to be performed without sufficient coordination with the SAM team. Having a strong sponsor in executive management will help manage potential resourcing and priority conflicts as well as help define correct SAM processes across the company. The SAM sponsor would also have quick and direct access to other company executives if any urgent or strategic topics need to be discussed and decisions made. Without a senior sponsor, SAM may become an unimportant function that observes changes and fights fires reactively.
A strong sponsor may also help significantly with a very important step for the SAM program – to get attention. In large companies, it always takes time for executive management directions and strategies to be understood and accepted by all (relevant) people in every organizational unit. A lack of understanding of needs and goals can lead to delays and inefficiencies when a SAM program tries to implement new solutions and processes into the community.
However, understanding the needs and goals doesn’t mean that people will start cooperating immediately or happily. Even well-defined program and project structures don’t ensure smooth cooperation and execution. People assigned to the project may have a lot of other work to do, may be afraid of the coming changes or may have very different perspectives caused by the local perspective (local and global perspectives and priorities may be very different at times). From my experience, ensuring cooperation across the company is the most critical success factor when it comes to global SAM.
Often, there are external parties involved in the SAM implementation, e.g. vendor experts or tools providers. Smooth cooperation with these parties is equally important for the success, of course.
When the right attention and cooperation are in place, then it’s much easier to achieve results. There will always be issues to overcome and well-organized and motivated people can deal with those issues better and faster. I also recommend planning the project in a way that has results in place quickly. If it takes too long to get initial results and benefits, people (in all levels and roles) may lose focus, interest or even hope. In other words, the initial scope or the first phase of a SAM implementation should focus on areas that aren’t too complicated or wide. Those can be done later when there is more experience and buy-in in place.
SAM is not a one-off activity and an implementation of a SAM program in a company is just the first step. It would be a real waste of effort and money if there is no sustainable SAM model in place before the SAM project is finished. Please make sure that implementation of business as usual (BAU) is one of the key deliveries of the project and that there is sufficient bandwidth, capabilities, budget, etc. in place for SAM execution and continuous enhancements. Even the strong sponsorship needs to stay in place for BAU in order to ensure that SAM retains the right attention and importance.
Everything changes and needs to be improved over the time – and it’s surely true for SAM. I recommend keeping the same “cycle” logic for continuous enhancements and change management with the result that things should go smoother and smoother over time with the growing experience and maturity.
Data is another key element for SAM success. In large and non-centralized companies, it may be a tough job to collect all necessary SAM data with high quality. However, wrong or incomplete data produces wrong results which may significantly impact overall benefits from the SAM implementation and put the whole program in question. Both technical and business data may become a challenge, but technical data is typically more complicated in the long-term and I would concentrate on that data.
Driven by your company size and complexity, you may have a few data sources delivering SAM data or have hundreds. You may want to deliver data directly from these sources into a global SAM tool or implement multiple layers, e.g. data source -> CMDB ->SAM tool. You may have sophisticated and automated scanning tools in your data centers or may be collecting data manually via spreadsheets. In all cases, you need to pay high attention to data completeness and quality. I recommend the following logic be applied to SAM data:
Before data collection really starts, you should understand what data you need and how the data can be collected. The typical questions would be: Where are all of my devices located and who owns them organizationally? Do I have all of the hardware attributes necessary for SAM collected? What is the impact of virtualization? Should I collect both installations and usage? Do I need to collect some specific information for products that are in scope?
If you don’t know enough to answer questions such as these, you may involve (internal or external) license experts who should be able to help. Starting with data collection before such questions are answered would probably lead to surprises, rework and dissatisfaction.
If you know what needs to be collected, you should define how to do it. You will find some scanning tools on the market which may help a lot. However I suggest you evaluate their capabilities against your scope and specific needs first. I also recommend proof-of-concepts or pilots to confirm those tools would really deliver to your needs and expectations – before any major investments and rollouts. Sometimes, you may even find out that manual data collection works the best, e.g. for small/special estates with less frequent changes. Of course, conducting a manual inventory requires both effort and discipline which may become a challenge over the time. Significant attention also needs to be directed at the sustainability of the collection. Data has to be collected regularly and complicated/heavy collection processes may create issues long term.
My experience is that you should never trust collected data, especially at the beginning. Validation of data completeness and quality is absolutely critical step both during the SAM implementation project and also with business as usual. With a large and complex infrastructure there is always a high probability of data quality issues even if the best available tools used. Data validation is typically a disliked activity and if ignored or underestimated, the issues return, possibly at the worst point in time. I’ve also learned that data must be validated by people who understand its meaning and can also manage the improvements. In other words – outsourcing of data validation simply doesn’t work well.
I recommend you process validated data only, e.g. you calculate license balance using data that was checked before. There is surely an option to upload all available data into your SAM tool and do the first validation in the tool directly, but I see potential for a significant risk when this approach is chosen – initial results coming from the tool would be typically very wrong and people lose trust in the tool even if the real problem lies elsewhere. An untrusted tool can become a very big obstacle on the SAM journey and it may take a lot of time and effort to get that trust back.
It may be impossible to validate every single piece of data ahead of use. In reality, there will be some additional validations steps triggered by the results; it is only when a license balance is created that some quality problems become visible.
Results from a SAM tool often trigger the next discussions related to data completeness and quality. The visibility of results is the very major step in SAM – what becomes visible and measurable can be managed efficiently. After results are published, it is time to review the whole cycle again, typically finding smaller and smaller gaps and surprises over the time. I strongly recommend that you declare your SAM tool as the “single version of truth” and make all efforts to ensure the results from the tool are correct and accepted. Building or keeping any alternative tools, spreadsheets etc. “until the SAM tool is ready” would just slow down the implementation and introduce confusions (which results are correct and to be used?).
I still see SAM as relatively young and immature business practices. With more and more experience and maturity, it will become simpler and simpler over the time…and one day my friends and colleagues will be right saying that SAM is simple!
Of course software manufacturers themselves could make our SAM lives much easier but that would be another story…