Cloud computing’s impact on mid-sized retailers

- April 12, 2010

Santa Clara, CA – April 12, 2010 –

At this point in the evolution of retail cloud computing, we remain steadfastly in the early-adopter phase. Many companies see virtualization as a potential for flexibility, scalability and reduced costs, yet most have not yet taken the plunge into the cloud.

When setting up the server environment today, most providers will ask the retailer for the load during peak times throughout the year. For many retailers, 90%-95% of their traffic comes during the holiday season, leaving them to pay for that same level of support the additional 10 months -- regardless of whether that bandwidth is actually being used.

Cloud computing and virtualization have the ability to change this model. Retailers can put software into the cloud, and rent a server for 10 cents an hour. Companies can scale up (or down) based on customer demand, holiday timing, and general growth.

Many larger retailers are reticent to move to the cloud, mainly because the thought of entrusting a billion dollar channel to a virtual environment makes them nervous. Couple that with PCI compliance worries, gargantuan amounts of customer data, and large, sometimes slow-moving infrastructure, and big retailers will likely be the last to adopt this new model in coming years, if ever.

However, while cloud computing does offer potential for any size retailer, big or small, mid-sized retailers in particular seem best poised to take advantage of what the environment has to offer.

1. Can react quickly: Smaller technical teams and tighter decision-making circles allow mid-sized retailers to evaluate a new opportunity and put a plan into action faster than their larger counterparts.
2. Have more manageable amounts of product and customer data: Large, big-box chains deal with more SKU assortments and a broad population of customers. Mid-sized specialty retailers have a smaller store footprint and associated SKU assortment and tend to have a more loyal/dedicated following from their customer base.
3. May have greater gaps in peak versus slow times: General merchandise/big-box retailers sell continually throughout the year.  Specialty and mid-sized retailers tend to have a very seasonal business, and gifting plays a large role in their sales.
4. Have fewer backend systems to integrate with: Large retailers will often purchase different, best-of-breed systems for inventory management, fulfillment, merchandising, etc. This makes integration more complicated and costly. Mid-sized retailers will tend to buy a common platform to run their business and will have fewer integration points.

Beneficial scenarios for cloud computing
In addition to the criteria outlined above, mid-sized retailers are often faced with the following scenarios, which lend themselves naturally to the cloud computing environment.

1. Geography: Retailers operating in Europe and Asia can operate clouds all over the world, without having to involve outside resources. Once established, a company can spin up additional servers on their own with relative ease. Cloud computing can also be an ideal fit in supporting the heavily mobile cultures in Asia.
2. Clearly defined peak volume: As mentioned earlier, many retailers do 90%-95% of their business during the holidays. Yet all retailers don’t necessarily fit this model. Specialty retailers may have a one or two-month window tied to seasonal activity, and so the peak could come in the month of May. The important factor is if the retailer can clearly define this window consistently year after year, and react appropriately.
3. Viral brand: A retailer relying heavily on viral marketing to launch a new promotion or product can capitalize on the flexibility of the cloud environment. The company can react to the unpredictable demand generated by a campaign without heavy investment up front, yet can rapidly expand the environment should the campaign be successful.
4. PCI attention: Operating in a cloud environment forces a retailer to focus on PCI. In a shared virtual environment it becomes more difficult to meet the PCI compliance standards knowing that the retailer has little control over the computing environment. Retailers in a cloud environment can use tokenization techniques to securely store credit cards off site with a trusted third party.

One of the largest impediments to cloud computing remains the traditional perpetual license model that software companies try and pursue in a cloud environment. The whole purpose of the cloud model is to provide utility (pay-as-you-go) computing. Software vendors will need to adapt their licensing to be usage-based. The cloud computing infrastructures provide the billing mechanisms to record usage as servers are turned on and off.  Large-scale retailers will most like go with private or internal clouds to maintain control over their applications.

Similar to the evolution of the Saas model, cloud computing will offer retailers scalability and flexibility, but may take several years for companies and service providers to work out the appropriate models. Mid-sized retailers will remain well-poised to take immediate advantage of this new technology in the months and years ahead.