Thanksgiving – a full plate of SAP data
With our Thanksgiving plates soon to be overloaded with turkey, candied yams, honey glazed carrots and pumpkin pie I start thinking about the aftermath.
Black Friday and Cyber Monday (we couldn't think of a better name?!) will be on us while our bellies are still full and round and retail businesses of the country will have warehouses stocked to similar levels.
Retailers are operating in a world which has changed enormously over the last few years; the information available immediately to the smartphones and tablets has empowered the shopper.
Operating as efficiently as possible to meet their demands, being able to innovate and implement changes rapidly to keep your place as the first-click.
For this week and last, year on year, Amazon showed a 13.9% and a 14.6% growth. With 44% of that coming from phone traffic it is clear that the mobile eComm market continues to drive sales.
On their own sales day in July they were making 398 sales a second - 34.4 million transactions. That is a LOT of data. It is no wonder they own their own data centres!
Big SAP Data
Every SAP system will contain a massive volume of data. From ECC to BW, CRM to SRM. Each transaction creates a few more bytes.
Data costs money to store.
It mounts up, just like your thanksgiving plate.
A few terabytes here and there, a couple more tater tots and soon it will be an indigestible cost.
This is without even talking about mergers and acquisitions, and the associated consolidation of ERP systems.
Yes, it is important to keep that data around for reporting, analytics and sometimes for legal purposes too but with all this going through, data volumes are going to increase massively and system performance is going to be reduced. Just 4 real world scenarios which will benefit from archiving.
Potential website customers are going to get frustrated waiting for the page of new lines to load because you haven’t archived your old stock items or assortment lists.
Order processing speeds are degraded because 5 year old documents which are not relevant are slowing up the searches on the database.
POS terminals are either out of date or slow because your daily assortment list update takes so long.
Your forecasting and replenishment tasks are not as frequent as you like because the overnight jobs have to wade through old inactive or completed items. They might not have a bearing on current rescheduling but they still count to database runtimes.
Archive all your unwanted data
If you take those millions of transactions from previous years, not just from the holiday period, you are probably running some form of archiving to put aside those records which don't need to be instant-access, but can be part of your reporting layer.
Remember, archived objects shouldn't be just gathering dust never to be looked at ever again. Your reports and your predictive analytics needs that archived data to deliver insight.
In the same way as I will take a nap and shut down all critical TV and conversation functions after my thanksgiving dinner, you will probably be reducing all non-critical jobs to allow customers an unrestricted free-run at the biggest shopping hoopla of the year.
Feel free to do the same with your archiving, but don’t forget to turn it back on. Or better still, automate and accelerate so that after a couple weeks of holding your breath you won’t be left gasping.