Article Series

Unified Commerce: Implications on Traditional Back Office Systems

Unified Commerce: Implications on Traditional Back Office Systems Part 8

May 8, 2017
   |   by 
Rick Boretsky

In this eight-part series, we’ve examined each of the cornerstone back office systems to reveal the issues in integrating them to the world of unified commerce. This blog focuses on the new challenges Unified Commerce places on Analytics. This will serve as a useful review to the reader of the previous 7 blogs, each of which identifies the challenges Unified Commerce brings to traditional back office systems.

Where does Analytics fit?

Most retailers employ a mixture of back office analytics platforms ranging from Excel to broad business intelligence systems from Microstrategy, SAP, or IBM. Excel is pervasive in retail because of its low cost and users can do their own manipulation. More and more we’ve noticed that retailers are turning to Excel of Unified Commerce Analytics, because their sophisticated Analytics platforms are not yet configured to:

• consolidate information across channels

• add enriched UC attributes

• analyze important data relationships

• create more insightful UC-oriented metrics

Excel is an amazingly powerful tool. But it introduces significant issues about data integrity, system administration, and user productivity. Our challenge as integration specialists is to ensure the integrity of essential source information to the Analytics Platform without introducing more complexity. However, if Excel is the tool our clients prefer to use, we do our best to fully automate the flow.

In the previous 7 blogs, we’ve traced the flow of data in a Unified Commerce environment and we noted where fundamental changes are needed to the back-office systems. We’ve seen that Unified Commerce challenges the apparent simplicity of data flow we inherited from the brick and mortar era. Now we must deal with much more.

But no matter how clever we become in enriching the old data flow with new attributes, the traditional back office systems will never be able to keep pace with demands of Unified Commerce for analytics. Eventually, you can work towards replacing these traditional back office systems. However, the retail software industry is toiling to provide comprehensive solutions that do not completely disrupt the organization. So, the smart strategy is to rely on a robust analytics platform that interacts with the current back-office systems to meet the growing needs.

A discussion about Analytics typically involves four elements: source data, data model and attributes, metrics, and aggregates.

Unifying “Path to Purchase” Source Data from digital commerce into traditional structures

The previous blogs have identified numerous new sources for transactional data that Unified Commerce retailers ought to consider in their Analytics. In summary, these new sources fall in four broad categories.

The data pertinent to the “path to purchase” establishes a foundation for management at all levels and helps provide insights into fast changing customer behavior. We recommend retailers try to collect the following “non-sales” data in their sales and order management data streams. The goal is to correlate this data with customers and, when appropriate, their transactions.

1. Geolocation of shopper

2. Click stream

3. Offer response

4. In store dwell time

5. In store, pickup arrival

6. In store digital advertising interaction

7. Product information request

8. Email response

9. Call center interaction

10. On line item returned

11. Salesperson appointment

12. On line customer reviews activity

Much of this information is considered “unstructured” data by industry observers. But our experience is there is always a structure which enables cross referencing a significant event with the key retail tables.

Digital marketing and customer order processing information, when well integrated with other information, provides insight into how Unified Commerce both erodes margin and provides uplift.

These data points include:

1. Labor for customer order fulfillment

2. Delivery costs

3. Customer accommodations

4. Discounts and coupons

5. Loyalty program incentives

6. Targeted marketing costs

7. BOPIS upselling

8. Sales commissions

9. Fees to digital commerce facilitators

10. On line and social advertising fees

11. Customer return processing costs

12. Double crediting of sales

Inventory data must be broken down to actionable elements and divided into unit and cost.

Inventory data includes:

1. On display

2. Reserved for BOPIS

3. Back room reserve

4. Available for sale

5. Recent customer returns

6. In Transit

7. Known shrinkage

Additionally, there are significant costs to be captured in the Logistics and Workforce Management data:

1. Labor handling costs

2. Shipping costs

3. Storage costs

4. Labor for customer order fulfillment

5. Warehouse shrinkage

6. Packaging cost

7. Cost of cycle counts to ensure accuracy

8. Cost of damage and returned good

Analytics Data Model and the shared enterprise tables.

In many ways, the strength of a retailer’s Analytics program depends the robust attribution of key tables like Product, Location, Customer, Inventory, and Sales. Unfortunately, since Unified Commerce retailers no longer run one-size-fits all operations, most of the traditional retail metrics inherited from the brick and mortar era no longer present themselves in an actionable form. For example, in the brick and mortar days, when COMP SALES was trending down, management would undertake corrective action in pricing, associate training, or promotion. Now, down trend on COMP SALE might indicate that digital commerce is particularly effective in that store’s immediate environs. Once doesn’t know until sales are segmented by type of customer and shopping behavior. This explains partially why so many users expend so much effort to create to their Excel spreadsheets. As we’ve said time and time again, it’s a data integration problem!!!

You can download our Unified Commerce Data Model Attributes Guide to identify some of the attributes which, once added, will provide more robust analysis.

Why the old metrics don’t and can’t provide insight

The primary purpose of Analytics is to provide enough insight about a problem or opportunity so that a user can take action. Analytics can only fulfill its promise when it correlates data with actionable retail metrics and it is summarized correctly. Often insights come in the form of anomalies or trends which, once revealed, promote an investigation into the supporting detail.

Unified Commerce demands the adaptation of every traditional metric, from COMP Sales to Inventory to Margin to Plan, to the Unified Commerce realities. As we’ve pointed our repeatedly, Data Integration is the key to providing meaningful Unified Commerce metrics.

You can download our Unified Commerce Metrics Guide to learn more.

Building the actionable Aggregates to measure performance by Customer Segment

Traditionally retailers looked at their performance information through the lens of their product and location hierarchies. Unfortunately, these perspectives mask the answers to the Fundamental Question on every Unified Commerce retailer’s mind: who is my customer and how is their shopping behavior changing? As we’ve tried to point out, the answers to these questions are entirely dependent on the data integration specialist job of supplying the underlying data model with information about customers and their shopping behavior. This vantage point overlays the traditional reporting hierarchies and enables actionable insight across the enterprise at every level of responsibility. Users from all over the organization can use this insight to modify the business to reflect a clear picture of today’s retail reality.

Conclusion

This concludes our blog series Unified Commerce: Implications on Traditional Back Office Systems where we’ve tried to elevate the importance of Data Integration in this fast-changing environment. We are passionate about this work at RIBA as the retail industry moves past the “all hands on deck” mode of working. Hopefully this series will help our readers understand the importance of giving data integration the attention it deserves.

We conclude with three overriding points:

1. Think architecturally about your data integration needs. Stop thinking about integration as an “all hands on deck” project.

2. Enrich, don’t limit, your data to accommodate Unified Commerce transactions.

3. Rely on Analytics to house and organize additional data flowing from Unified Commerce transactions.

Please let us know how if you’d like us to send you the eBook that brings together this entire series. And as always, let us hear from you. We’d love to hear how you are facing the data integration challenges we’ve discussed here.

  Previous ArticleNext Article   ⟶