In the fourth of the series of technical leaning blog entries, our EMEA Technical Lead makes the next stop on our tour: how organised data at the start of a project is key to its success. The data and methods typical for a Proximity Insight clienteling implementation are discussed. Barry refers to his personal experience and knowledge of being in Retail Technology for 30 years.
Read Barry’s first blogpost to learn about his extensive retail career and why he loves the art of clienteling.
Organised data is information & insight
‘The goal is to turn data into information, and information into insight’
Carly Fiorina – ex CEO of Hewlett Packard
I recently found the above quote from Carly Fiorina and it perfectly encapsulates one of the key aims in a successful CRM or clienteling implementation. In delivery of an implementation, the existing data in an organisation is integrated into Salesforce where the Proximity Insight clienteling application resides. With the data contained there, it can be turned into related information the store associates or head office users can see. From that, the insight is generated either by the person seeing the information in one place and taking action, or processes and reporting which can provide the insight. This then drives the key business decisions to generate and uplift sales through targeted personal outreach, virtual or in-store appointments, and customer specific microsites to name just a few examples.
Hence, obtaining the aim of organised data is a significant early step and can lead to quicker implementation of the product.
The Three Customer Data Cores
Clearly a source of customer data is needed, and already having a view or source of customer data is a huge advantage for a brand. The basic data includes a customer’s name, contact details and consent to communication, however any extra data or metrics around spend, product preferences etc adds even more value. From such information, Proximity Insight is able to run reporting on attribution, outreach effectiveness, spend analysis and more to leverage further insight. We also enable automated customer journeys that provide the users with tasks and suggested outreach for the customers.
The level or type of information to be imported or captured on a customer does need to be defined and there may be differences between what a user in store needs to see versus what a user in head office view as important. Through page layouts and profiles, Proximity Insight’s clienteling application can present the key pieces of information appropriate to the user to enable the insights.
One key consideration is not only the flow of data into the application but also where in the app new customers or updates should be posted. Without being too technical, mapping out the customer data journey is key in deciding not only where the data is coming from but also how to pass back changes to the source.
While mentioning a single source or view of the customer is preferred, it certainly does not prevent an implementation from proceeding. Using Salesforce can bring tools and processes to help create the single view of customers that has potential to be delivered to other areas of the business. In these situations, the customer data journey is extremely important as the single view may need to update the existing sources.
A key aim for the app is to enable product recommendations to be made to the customer and also provide a visual representation of previous purchases. As such, a product feed is required and can be as detailed as preferred to provide information to the user. Key pieces of information are product name, description (typically the long ecommerce description), size and colour, one or more image URLs, price and finally an identifier that transactions can be connected to. This can be extended with filters such as category, brand, and type amongst others.
One area that often needs to be reviewed is the level to group styles. This is especially the case for fashion brands. Typically in fashion, the grouping is either the style itself or the style/colour. How the product is grouped creates the parent record and then the child will be the individual sizes underneath. The reason for this distinction is for recommendations. The best approach is to work at style/parent level then the transaction data will link at the size/child level. This way, if the customer opts to buy a different size then the recommendation and attribution will not be affected.
While in theory only customer and transactional data is sufficient, being able to have a product feed drives recommendations, basket and customer bespoke microsites.
The final data core is the transactional data for customers and is the main link between the two described above. Ideally this will include historical data loaded at go-live, however the real driver is a daily update of new transactions that link the customers and product loaded into the platform.
The level of information to facilitate the transaction piece can be quite minimal, with the basics being customer, product and transaction IDs. However, being able to identify the source or location of the transaction along with (in the case of stores) who completed the transaction provides significant gain in insight. This also allows a level of gamification or attribution within store teams as it can show the conversion between recommendations to the products purchased.
Pulling It Together
Once the sources of data are identified then the process of integration can be defined and will vary depending on the systems and platforms. As Salesforce is widely used it’s possible those systems might already have integration to the platform and it is highly recommended those are explored.
Alternatively if there is an ETL or other data integration platform in use then it’s likely that it’ll have the required connectors. Where possible, the preference is to reuse what is in place today to speed up implementation. An example of we have had great success is where an existing feed is created for Google Shopping, as that has nearly all the required product data.
If no other option is available then Proximity Insight is able to leverage Talend and point to point API integrations to enable the data flow and help organise in to information and insight.
Overall, the quickest and easiest implementations have been where a clear view of the data and the requirements is set early on in the project. This will provide the organised data and insight.
Alternatively, going through the process of implementing the Proximity Insight clienteling platform will enable organising data into information and deliver the insight needed to grow sales and outreach.
Get In Touch
I hope this was of interest, and if there’s anything you would like to see covered in a future blog then get in touch. Thank you for taking the time to have a read.
By Barry White EMEA Technical Lead Proximity Insight