Multichannel Retailing Reference Architecture: Update

The Multichannel Retailing Reference Architecture

This is a minor update to the architecture after the comments on the original posting. I've now added the 'Decision Support' component.

This component should contain the enterprise-wide reporting and Business Intelligence (BI) tooling not encompassed in any of the other components, such as the CRM. Ideally all reporting and BI should be performed using this component, though the reality is that many application packages will provide their own specialist reporting tools that may be better suited to the job at hand. These tools may be best placed for operational reporting, whereas the tools within Decision Support will be for non-operational reporting, such as data mining historic information and trend analysis.

The challenge here will be deciding what should be considered as operational and non-operational data. One possible approach could be as follows.

Partition the 'Master Data' in to two areas; Operational and Historic. The Operational Data Store (ODS) may enforce a policy of containing only 13-months of data, whilst the Historic Data Store (HDS) may contain all older data. Of course, each entity in the ODS may have data of varying age due to legislation and compliance issues, much as data in the HDS will also be governed by such factors. From a cost perspective the HDS may go further and implement Information Lifecycle Management (ILM) techniques to store the data, e.g. 1-3yrs online, 3-7yrs automated/retrievable offline, 7+yrs offline/offsite.

The tools providing reporting on the HDS are always likely to be generic, it's more cost-effective than having secondary instances of each of the business applications for specialist reporting on the historic data. However, this is where there will always be the challenge of the business user requiring the same level of reporting they have on their operational data. Though this frustration can usually be circumvented by demonstrating the data mining and flexibility of the BI tools. The only time it may be necessary to provide secondary instance of a business application against the HDS is for legislative compliance. A recent example I have experienced was to ensure a despatch system could reproduce all labelling for any customer orders shipped internationally for a period upto 7 years. To achieve this, we had a secondary instance of the application configured against the HDS that would only be started upon business request (e.g. during an audit) and agreed with the application vendor that this would not constitute the need for an additional software licence. A cost-effective solution, satisfying both the business stakeholders and our IT budget!

Multichannel Retail Roundup

This is the first in what I hope to make a regular spot on the blog where I'll be rounding up and summarising articles, blog posts and generally any content that's caught my attention relating to Multichannel Retail. Hopefully you'll find some of it interesting and of relevance.

UK Retailers continue to invest in multichannel programmes
First up is this recent article documenting the multichannel round table debate hosted by BT Expedite. Looks like it was a very open debate, with some great soundbites captured and addressing some common issues across retailers. I particularly like Andrew Clarkes comment relating to TopShop not becoming a social networking site, which RetailWeek decided to highlight in a callout. Reading between the lines it sounds like there could be some disagreement on strategy there between the parent and operating company. Another notable mention in the article was Comet's click to chat, whilst they were reluctant to give out figures the article seemed to suggest they've seen a measured impact on conversions. As a slight side on this, an ATG partner recently implemented a callback feature for a Health Insurance client. An email was automatically sent to the call centre when a customer abandoned their application, the call agent contacted the customer to see if they could be of assistance - this resulted in a 17% reduction in abandonment, imagine something similar for an online retail site as part of checkout abandonment.

How a Company’s Culture Can Affect Efforts to Integrate Channels
This article published by Retail Systems Research provides an interesting insight in to the multichannel strategy of US retailer, Bare Escentuals. Clearly hampered by their US growth strategy, whilst complex was primarily product-centric and store based (typical of most traditional brick'n'mortar retailers world-wide), are using their EU and Asia expansion to implement their multichannel vision. The final objective to migrate their US operations to the multichannel model once it has proved itself. This is a relatively risk free approach, as it removes any potential impact on their currently successful, if slightly hamstrung, US business.

Blog: Mobile Retail - Coming to any store near you
Finally in this roundup, a fellow blogger's post on mobile retail and the recent Forrester Report. Some nice references to recent mobile retail applications and how consumers are now using mobile devices more and more to inform their buying decisions whilst on the move.

Crowdsourcing and Online Retail

Before leaving the house this morning I caught a TV news article about http://www.galaxyzoo.org, a new website asking the public to help classify star systems and spot oddities in the vast number of photographs scientists have of the universe. What a great use of crowdsourcing I thought, how could online retailers benefit from such concepts?

An obvious example might be a competition to design a new product, similar to Walkers recent efforts to get the public to define new flavours of crisps.

Possibly slightly less obvious would be to address a couple of issues most website managers would profess to having, search result relevancy and product hierarchy. So, how about having something similar to Googles new additions allowing your users to rank site search results
or comment? Or allowing users to re-categorise products? It would be an interesting experiment to see the affects on conversion rates. The theory being these would improve as the customers with common interests improve each others routes to find products.

Using Social Media to drive footfall and customer retention

I've just been reading Joe McKendrick's SOA blog over at ZDNet where he's been blogging from the Microsoft Fastforward conference in Las Vegas. He's described what he calls the LIFT phenomenon; Linkedin, Facebook and Twitter.

I was thinking, with a good Social Media marketing strategy in place a retailer could really benefit from these platforms, gaining insight, introducing rich interaction with their followings.

Many retailers have RSS enabled their websites, but what if it were just as easy on their eCommerce platform or Enterprise Marketing Mgmt (EMM) tool to click a box when creating that latest and greatest promotion and it to appear directly on those Social Media sites?

With the advent of FaceBook Connect and the Twitter APIs that shouldn't be so difficult should it?

If these customers have already pledged allegiance by following on Twitter or becoming a Fan on Facebook, imagine how special they'll feel if they're always the first to know about special offers and discounts.

Multichannel Poster Child: Argos

Here in the UK one of the retailers regularly held aloft for being a successful multichannel retailer is Argos. Pioneering kiosks, click & collect amongst others.

However, compared to most store based retailers, Argos have a couple of distinct advantages:

• Each store is effectively a mini warehouse with tight stock control
• The customer journey is almost identical across all channels

Let's take the latter point and observe the typical in-store customer journey...

1. Customer enters the store and is presented a catalogue to browse through
2. Products of interest are placed on a list and can be checked for availability
3. Customer takes the list to the checkout and pays
4. They now wait for the product to be picked from the local warehouse and delivered to the collection desk in the store

Does that all sound familiar to those of us working on eCommerce websites? How easy must that have been to transfer this business model to the web and phone ordering? "Hey Joe, this Internet thing, how do we get ourselves on it? - Heck! It's just a virtual version of our store with a large warehouse attached, we just need to get the home delivery logistics in place"

Having only display stock on the shelves means they must have much more accurate stock visibility, none of this late check in of received stock only for it to already be out on the floor and half of it already gone! (of course this wouldn't be issue if there was real-time reconciliation between ePOS and stock management systems, as well as automated check-in).

Such stock accuracy means cross-channel functionality such as 'Click & Collect' can be implemented with confidence.