The cost of downtime for retailers is increasing
The typical cost categories of downtime are lost revenue, mitigation, lost productivity, stock devaluation, and brand damage. If a retailer manufactures their own items, then you have the added costs of lost production, lost capacity, inventory carrying costs, impacted customer responsiveness, and direct labor.
Unique challenges in the retail sector
The retail industry has some unique pain points; some are born from the nature of the business itself, while others are the result of a global shift in how, when, why, and where consumers shop. What was already in a state of transformation became accelerated with the onset of the global pandemic.
Here are some downtime statistics specific to the retail industry provided by Digi and LogicMonitor:
- 81% of retailers experience downtime at least once a year.
- 87% of retailers wait up to 4 hours for support when an outage occurs.
- Even with so-called 99% uptime, businesses will still experience 80+ hours of unplanned downtime per year.
- 68% of retail sector companies surveyed felt that an outage or brownout could happen at any time, and that such an event would attract national media coverage.
During periods of downtime, retailers experience:
- Lost sales
- An inability to process credit cards
- An inability to access customer purchasing history
- An inability to locate items in stock, at a distribution center, or at another retail location
- An increased vulnerability to data breaches
- Brand damage
- Frustrated customers
The last bullet is particularly concerning. Despite all the costs and negative impacts of an outage or brownout, and even with IT controls in place, companies still feel vulnerable to such events. LogicMonitor found that 51% of companies surveyed believe outages are avoidable. Conversely, that means 49% of companies believe outages are unavoidable.
Changing times means different challenges
In the days of brick-and-mortar, downtime causes for retailers were the same as any other industry: network failure, infrastructure failure, software issues, third-party provider outages, and human error. The Standish Group calculates that POS outages cost the average retailer 4,700 USD per minute, or 282,000 USD per hour at retail locations. Customers do not like waiting in line, particularly in the United States where customers on average become frustrated after standing in line for more than 2.5 minutes if there is no progress. 1 in 3 of those people will abandon the line altogether after 5 minutes.
The concept of waiting has not subsided with the advent and rapid growth of e-commerce. In fact, our collective patience has become thinner. Americans, for example, have begun to alter their brand purchases in light of recent global events. Communications company Ketchum puts this number at 45%, while McKinsey puts the number closer to 75%. In light of this, outages and brownouts now have a greater impact on consumers who simply won’t tolerate delays or other impediments to consumption. There are always other options.
Online sales processing is critical
According to Census.gov, 2020 e-commerce sales totaled 791.7 billion USD, or 14% of total US sales. This is in line with global numbers which indicate 4.1 trillion USD of 2020 retail sales, of which 14.6% were transacted online. The combination of a growing percentage of younger consumers and COVID-19 has pushed online sales upward. Many retail stores have seen a huge increase in online order fulfilment volume. An August 2020 article by Digital Commerce noted that Home Depot’s stores fulfill 60% of orders online.
This means that downtime isn’t just a retail location inconvenience (you could always hand the cashier cash vs. processing a credit card). It can halt business altogether if an outage impacts both POS systems and online ordering. Now atop the list of outage sources are usage spikes (i.e. too many people requesting online services at once) and DDOS attacks. DDOS attacks have skyrocketed in recent years, with an estimated 10 million occurring in 2020. Worse, these attacks are increasingly multi-vector, up some 2,815% in 2020 from 2017.
The cost of an outage can be painful. The online retail behemoth Amazon lost an estimated 34 million USD per hour in 2018 because of their well-publicized outage. Imagine what their loss would have been if that outage had occurred at the height of the pandemic?
So, if unplanned outages cause such damage, it behooves companies to keep planned downtime to an absolute minimum. Any major IT transformation project, such as a migration from legacy SAP to Suite on HANA or S/4HANA, will require some period of downtime, but this can be minimized with a combination of advanced technologies, best practices, and basic cutover planning.
For the retailer, particularly one that is continental or global, there is no good time for downtime. Once sacred days like Christmas, New Year’s, and Thanksgiving are now business-as-usual in many parts of the world, and when one retail location is closed for the evening, another is bustling with activity a few time zones away. And of course, e-commerce never sleeps.
Keep downtime to a minimum
In light of this, how do you keep planned downtime to a minimum? Using the aforementioned scenario of migrating to S/4HANA, I suggest the following:
- Know your system inside and out. This may seem like a no-brainer, but after years (or decades) of use and development, it is easy to lose track of what is being used, by whom, when, and how. Facts are your friends.
- Identify the best days and periods of time (or rather, least worst) when downtime could be withstood. Then, back up from these points to determine feasible project start dates based upon the scope and complexity.
- Take a page from the agile methodology and adopt an MVP approach to the migration. Ask yourself what you must migrate now (to have a fully operational and compliant S/4HANA system) and what you can migrate later. If you have a 10 TB system and only 0.5 TB is current and prior year data, migrate the 0.5 TB first, which will be far faster than the full 10 TB.
- Create a solid cutover plan.
- Invoke “freeze” periods when changes to master data or hierarchy data are not allowed.
- Give the infrastructure of your source system(s) a tune-up.
- Know what an hour of downtime costs your company and factor that into your decision-making process when selecting an approach or migration partner. A cheaper price when coupled with extended downtime could end up being far more expensive.
Employ technologies that:
- Automate the migration, validation, reconciliation, and testing processes.
- Can combine multiple prerequisite projects and activities into a single go-live.
- Offer deep insights into your system(s) and suggestions for optimization.
- Minimize throw-away work, applying a lean approach to project processes.
- Are geared toward the applications in scope – in this case SAP.
Retailers will be surprised how much the downtime window can be shrunk, but it requires planning and evaluating the options available to you.