advanced analytics

Advanced inventory analytics as a service

Do you find it difficult to really understand what your inventory data is telling you, or what you should do about it? Do you have optimization tools that are laborious to use or which give results you know to be wrong, but you're not sure why?

Leading organisations not only have visibility of what inventories they hold where, but also take advantage of sophisticated optimization techniques based on the scientific analysis of their empirical data. This gives them a clear advantage in making sure they have the inventory they need to serve their clients, while minimising excess and obsolete inventory.

Advanced analysis of large business data sets is fundamental to everything we do. We have developed proprietary technology that we leverage to provide clients with rapid actionable insights into their inventory data.

We help clients understand what their data means and how they can use it to optimize their supply chain through the application of advanced statistical techniques. By delivering analytics as a service we can be flexible and provide clients with insights that are most valuable to their circumstances.

Inventory evaluation

Our starting point with most clients seeking to optimize their supply chains is to carry out a diagnostic evaluation of their inventories. Using just 3 standard reports from your ERP system, we transform empirical data and run a range of algorithms to quantify improvement potential item by item as well as overall. We can do this in a matter of minutes even for multi-million-line data sets. The output is both an immediate high-level quantification of improvement potential and the basis of a road map to deliver quick wins and sustainable improvements.
nVentic inventory analytics

nVentic evaluations allow the quantification of optimization potential both overall and by item, with multiple segmentation criteria

Our evaluations allow the segmentation and prioritisation of all items of stock based on multiple criteria, including maximum reduction potential and highest shortage risk. They highlight slow and obsolete stock, provide parameters for replenishment control and collate all data in one easy-to-manipulate table.

Since our evaluations work on raw data from ERP systems, and since that data is usually of mixed completeness and accuracy, we set standard minimum values for some key parameters and provide clients with statistical analysis of the data quality itself. This allows clients to identify and prioritise data improvement initiatives. Our evaluation also contains flags that allow clients to identify items with particular qualities, such as declining demand or sporadicity, that require separate treatment.

Our evaluations provide an extremely rich set of insights that can typically generate several months’ worth of analysis and initiatives. They work as standard on empirical, historical data, but work equally well using forecast data or even a mix. They can also be used to model different scenarios and quantify the impact of changes to key parameters, such as lead times or service levels.

As an absolute minimum we spend half a day with clients presenting the findings of an evaluation, tailoring the session to their particular requirements and opportunities, so that they have a clear set of next steps to take advantage of the evaluation and turn the insights into action. Clients usually come out of these sessions with a greatly enhanced understanding of their optimization tools and how to deal with non-standard items.

While every data set is unique, our experience over multiple years, organisations and industries is that despite the very conservative approach we take to quantifying potential, the improvement identified by our evaluations is almost never below 20% and is not infrequently 40% or more. The subsequent projects also show the figures to be very realistic, even though it may take more than a year to realise the potential identified in full.

Advanced supply chain analytics

Our evaluations themselves are highly advanced and generate multiple insights, combining a range of analyses. In addition, we have automated a number of supplementary routines, to provide deeper insights into a variety of parameters, including lead times, order sizes and cover. Over and above these, we have also developed routines to look in more detail at a broad range of related topics.

nVentic’s advanced analytical capabilities cover numerous supply chain concepts, including:

  • Forecast accuracy, using a variety of methods
  • Service levels, including translation between different metrics (eg between fill rate, cycle service level, back orders and out of stock days)
  • Safety stock, using and comparing a variety of methods. The limitations of the “standard” formula are often under-appreciated
  • Back order analysis allows organisations to really optimize service levels, but back order data is often neglected, if tracked at all
  • Multi-echelon optimization. This is a concept which tends to be misused, and is frequently misrepresented. Our perspective is that it is preferable to optimize all echelons on a single echelon basis before looking at true multi-echelon optimization, which we support
  • Distribution of data (supporting multiple distributions, not just normal, along with the ability to auto-select)
  • Advanced simulations
This is far from being an exhaustive list.

All of these techniques are only as valuable as the actions they enable. The vast majority of organisations will get maximum value from our standard evaluations, but we have the ability to go much further with more advanced practitioners.

Technology and analytical approach

While we are not selling or licensing software, our technology is a clear differentiator that allows us to carry out consistent analyses almost instantaneously that others find long and laborious, if not impossible. We therefore like to be transparent about what we use and what clients get. Over 20 years, we have developed algorithms that apply the best and most applicable academic work on the mathematics and statistics of inventory science to large business data sets.

Our core routines have been coded and automated to remove manual work and the associated potential for human error, but we hand over the analytical outputs to our clients in standardised Excel formats. This means clients do not have to familiarise themselves with an additional user interface and have full flexibility to manipulate the data further themselves. It also makes it easy for clients to quickly carry out validity checks on their data and bulk upload improved versions back into their ERP as required.

Unlike “black box” solutions, we are happy to provide clients with full transparency into the methods we use. Indeed, we believe it is essential for supply chain practitioners to have this level of confidence if they are to trust the analytical outputs and use them in their professional work.

nVentic does not sell or licence its software for two main and interrelated reasons:
  1. Most organisations today already have some kind of optimization functionality in their software, whether within their ERP system itself or through other advanced planning tools. And yet we find the vast majority of organisations are using this functionality only partially, if at all. We would rather help clients get more value from their existing tools than provide them with another option for which they have to buy licenses and which then remains largely unused. Which leads us to the second reason:
  2. Our tools, like all inventory optimization tools, rely on a number of assumptions that will be correct to a greater or lesser extent for different items, are highly sensitive to a variety of parameters, and are difficult to understand for all but the most advanced practitioners. By providing our analytics as a service, we are able to help clients understand the limitations of their software, to identify and implement pragmatic mitigations of those limitations, and all in a format that is extrinsic to their day to day planning tools. We do not ask clients to put their faith in a “black box”, but seek to take them with us in understanding what their data means.

Would you like to talk to one of our experts? Contact us

© 2020 nVentic GmbH