Big Data Consulting

Fully packaged data analytics software is the core business of Lokad. However, every business is unique. Frequently the full potential of our technology and know-how can only be delivered through a certain level of customization, and in some cases it takes a whole new approach to best address the specific problem.

Software, not PowerPoint

The deliverable of most missions is a customized and implemented software application that solves an operating problem for your business. We are not interested in delivering man-hours and PowerPoint slides. Every mission starts with a clear definition of the problem as well as the criteria used to quantify the solution's performance.

Open source projects

High impact, high complexity business problems

We don't do data analysis for its own sake. Instead, we focus on the following situation:
  1. A clearly defined business problem or project goal
  2. Predictive analytics exploiting (large amounts of) data can solve this problem
  3. A direct impact on profits assures a strong ROI

Often, the problems we address can be expressed in simple questions:
  • What is the optimal price for an item?
  • What is the most profitable inventory level?
  • How much does a stock-out cost?
  • What is the ROI of a promotion?
  • What is the potential for a product we don't sell?
  • ...

Retail is the core focus of Lokad, however, we have completed successful projects in biotech, banking, energy and telecom.

Data agility, high-dimensional statistics and acumen

Big data agility refers to the speed and cost of manipulating large amounts of data. Many companies claim to be capable of transferring, storing and cleaning large amount of data. The question is: At what cost? A combination of low cost / high tech data analytics that leverage cloud computing allow us to keep IT cost very low.

High-dimensional statistics is often the key to turning data into really smart solutions. Producing a wall of metrics out of any dataset is both easy and mostly useless since employee time is the scarcest resource of all. The best numbers are the ones that nobody ever needs to read.

Business acumen is essential. We have gathered a team of entrepreneur-minded experts who understand the big picture. Furthermore, consulting is only a complement to our core business: We are not afraid to say 'no' when our initial assessment indicates that we won't create enough value.


Per Project Pricing

Usually, we do not charge man-days. Instead, we quote a project price for the deliverable of a milestone, which frequently includes a working software application. By delivering custom SaaS solutions, Lokad takes full responsibility for the success of the project.

Daily rates

  • Partner Data Architect: € 4900/day
  • Senior Data Engineer: € 2700/day
  • Data Engineer: € 1500/day

Early data discovery

Most of our projects rely heavily on the analysis of your data, and experience shows that an initial, rapid discovery of the data will increase the accuracy of project plan and timing. We therefore typically proceed with
  • an open business discussion, to grasp the pain points,
  • a data extraction, followed by a rapid data discovery analysis.

  • This will allow us to verify the feasibility of the initial project goals and in some cases result in the discovery of alternative goals, for examples when we identify unexpected 'low hanging fruits'.

    We frequently meet our clients in the initial phases of a project. Subsequently most of the work is done remotely. For productivity and security we only operates in our proprietary cloud computing environment.

    Areas of expertise

    Demand forecasting

    We are specialists for demand forecasting in large retail, wholesale and manufacturing. We know the particular challenges of these industries, and we have developed a strong technology and a set of solutions that address these problems. Some examples include:

    • Slow rotation products/long tail of the product portfolio. A classic problem for eCommerce, retail (particulary on the point of sale) and hardware wholesale is sparse data, i.e. products selling only infrequently. Classic ‘mean’ forecasting theory is unsuited for this type of data, and we have worked extensively on solving the problem. Quantile forecasting is our solution to this challenge. It has been known in statistical research for decades, yet we are the first to recognize its value for inventory optimization.

    • Point of sales forecasting for large retail networks. On the point of sale, a high complexity is introduced both by the sheer amount of data and by the sparsity of the data. Today, only very few retailers have sophisticated forecasts in place at this level, even though often more than half of a retailer's inventory is in the stores. We have been fortunate to work with Europe's leading retailers in this domain and developed point-of-sale forecasting technology that is particularly suited for the characteristics of store replenishment.

    Inventory optimization

    While working with our clients we have come to realize that forecasting is only part of the challenge, the translation into a smart inventory policy requires analytic tools and processes that optimize the reorder point and order quantities.

    Optimizing inventory is ultimately not only an operative but also a financial problem of balancing the cost of overstock with the cost of stock-outs. Building blocks include reorder points, order quantity and replenishment policy.

    Working capital optimization

    Inventory can tie up a lot of cash. The good news is that the potential for optimization in most businesses is significant. When exploited smartly, this can turn from headache into a strong competitive advantage. Exemplary projects include:

    • "Return on Inventory" optimization: Maximizing availability and turnover for a given amount of working capital that can be invested in inventory, or alternatively the release of working capital for a given service level.
    • Stocking decision: Especially in the long tail (low rotation products) the decision to stock a product or rely on cross-docking or drop-shipping is a difficult trade-off that requires reliable forecasting of sparse demand in combination with a thorough analyzes of the underlying economics.

    Pricing optimization

    We can help you optimize the pricing of product portfolio. Whether it is a full price store, promotional pricing or flash sales. As is the case for all of our software solutions, a thorough analysis of your sales data can provide an incredibly rich outcome in terms of revenue and profitability optimization.

    Loyalty card data

    Customer loyalty data is incredibly rich on information, and making sense of it is one of the most profitable analysis we have seen. Customer loyalty initiatives, promotions, cross selling, vouchers etc. are only a few initiatives that build on loyalty data.

    Out-of-shelf monitoring

    Grown out of our inventory optimization practice, we have developed a strong expertise in on-shelf availability. Our solution Shelfcheck has strong performance characteristics, and we have been able to build a lot of know-how in this area.

    What experts are saying

    Andreas Köpf A couple of weeks ago I stumbled upon the lokad-cloud open source project and finally found the time to review the source code. I am completely astonished by the solid quality of this library.

    @arwakatungu Loving @lokad. Makes #azure cloud storage so easy n code much cleaner

    @TechBubble Quite impressed with the @Lokad Cloud object-to-cloud mapper for #Azure http://

    @roblyndon What's the best thing to do after a heavy NYE night? Get to grips with lokad.cqrs of course, dummy.

    Chris Martin. I can't thank Lokad enough for the work you guys have put in. You have literally saved our startup thousands of man-hours!

    @browniepoints Check out Lokad.Cloud it's a service bus for cloud-hosted applications (they won Azure Partner of the year)

    @henrikwh I got to take a closer look at #lokadcloud. Looks like must have patterns for #Azure. Open source.

    @tomconte Keep an eye on this project, v2.0 is near! Lokad.CQRS for Windows Azure: Roadmap #CQRS /cc @abdullin @lokad