Data Analyst at 2 Sisters Food Group

In June 2014 I began my role as a data analyst for the technical services department of 2 Sisters Food Group. This role was as an assistant to the data manager in that department.

I began with making adhoc information requests from the over 30 factories that make up the company. This provided an opportunity to make contacts at each site and build up a rapport to aid all future work. I consider this to be highly important, since in a role such as this almost all communication with sites will be giving those contacts extra work and this needs to be remembered at all times.

I was then tasked with reworking the main reporting workbook (MS Excel). This held data on factory technical performance and was used to produce weekly reports as well as serve other non-routine information requests. A very short timescale was given when this was originally developed, and I was asked to streamline several elements to increase efficiency and reduce manual work. The main database also needed restructuring to be able to handle multiple years of data. Due to the diverse nature of the hardware around the business, all development was limited to maintain Excel 2003 compatibility, which presented several challenges (e.g. being unable to use SUMIFS() or COUNTIFS()).

Through the latter half of the year I developed my Excel and VBA skills on a number of projects for subsets of the business – mainly work on using Excel to automate and add value to existing reporting processes. I worked closely with my manager to learn how projects like these are managed from start to finish, as well as predicting timescales for complex tasks. I was able to visit a number of sites to gain an insight into how they processed the myriad of information produced within the food manufacturing industry.

Using what I had learned working on the main technical reporting process, I developed an environmental reporting process, together with a dedicated workbook, from scratch. It was very satisfying to be able to identify any difficulties that had been had with the technical system and avoid them ever occurring with the environmental system. This system has continued to develop and has become a core part of assessing the company’s progress in its sustainability plan.

Towards the end of 2014 I began experimenting with the improved functionality that would be available if we no longer needed to maintain Excel 2003 compatibility for the technical reporting workbook. This led to a system where every report ran from a central control panel, where users could view figures for any specified time range (with presets from a variety of retailer calendars), by site/division/whole company, and a wide variety of other settings made possible by the much larger number of conditions available within SUMIFS() and COUNTIFS(). Once we had assessed that all users of the workbook were no longer using Excel 2003 the new system was put into use.

In 2015 the challenge was to develop a way of scoring site technical performance based on a large number of KPI’s being assessed against multiple thresholds. With such a large amount of information to process, and a short window in which to produce reports following the submission of site KPI’s, this would only be possible if the process was automated. Automating the system also turned out to be highly useful as the thresholds we used regularly changed while the system was tweaked during development, and it was very easy to change them whenever required. It was during this development that calculation time became an apparent problem, so I also created a way of using the central control panel to selectively disable calculation for report worksheets that were not of interest.

When it was requested that the entire technical reporting workbook, together with a number of other pieces of information, be published to contacts at every site, it became a very taxing operation to publish everything in an accurate and timely manner. The next step was therefore to automate the production of a variety of reports, all under the appropriate settings, then automate production of the emails as well. This saved time and eliminated mistakes, and equally importantly reduced the pressure of the reporting routine on the person performing it. With reduced pressure comes the ability to respond to the unexpected (e.g. a request for extra information) and to more completely validate the information that is being produced.