How I Manage Successful Customer 360 Projects

Have you read about “Customer 360” but still aren’t sure what it is? Or maybe you’ve started on your Customer 360 journey but have hit some roadblocks? “Customer 360” is one of those nebulous terms that has different meanings to different people.

“Explain Customer 360 to me in a nutshell”

Ask 10 people what “customer 360” means and you’ll get 11 answers.

Customer 360 is defined differently by everyone. This is probably a good indicator that there is more marketing fluff to this term than real substance.
–dave

Customer 360, aka C360, quite simply, is interacting with your customers across multiple channels. The multiple channels could be social media properties and your own website. But Customer 360 is also gathering all of that multi-channel customer data and melding it with all of your internal applications that also house valuable customer data so you can use it at the time of the customer interaction to improve the customer experience.

Don’t you hate it when you call a company and have to enter your “customer id” to an Interactive Voice Response (IVR) system? And then when you finally talk to a human you have to tell them your “customer id” again. Then after 3 minutes of explaining your problem to the customer service rep she says, “let me put you on hold while I bring up your information on another system.”

If you aren’t sure what your Customer 360 project should solve first…start with this!! Tackling this first is foundational to any other problems you may want to solve.

Customer 360 is:

• a holistic, real-time view of your customer, at the moment of customer interaction
• crosses all products, devices, and systems. If there is a “customer interaction channel” it should be reflected in the C360 system.
• a system that delivers a consistent, personalized, and context-specific experience for your customer.

I’m a healthcare company, I don’t have a need for Customer 360

“360 degree view” projects are not only useful for gaining customer insights.

• Hospitals are building “Patient 360” systems to:
  • “Show me every interaction with my patient, from what in-network doctors they use to how old their receivables are, at the time they call me to schedule an appt”
• Product 360
  • Companies like to see what products are trending on social media, what the inventory looks like, and what stores need rapid shelf replenishment.
• Vendor 360
  • Manufacturers like to see dashboards that show them which vendors supply critical products that are late to delivery causing just-in-time manufacturing delays that are hurting sales when another vendor is slightly more expensive but has never missed a deadline.

Would you know where to start a Customer 360 project if you were asked to start one tomorrow? Are you convinced your existing Customer 360 project will succeed?

There are hundreds of firms that will help you with your Customer 360 journey. They will tell you why you need a Customer 360 project. And of course they’ll provide you with a catalog of services to help you. They’ll then tell you these are 12-18 month projects. There’s way too much risk with a long-term project like that.

And most of these vendors will NOT show you how they actually implement Customer 360. Even if you ask them pointedly. The vendors that do show you the HOWS want you locked-in to their products. Said differently, they are selling you a generic product providing them with recurring revenue. This might be a good solution, but you won’t know up-front.

I handle these projects differently, let me show you.

I don’t tell you WHAT you should do for YOUR Customer 360 project because I know every client’s need is different. I learn your problem first. And if you aren’t sure what your “problem” is…I’ll show a great place to start your Customer 360 project that actually shows value, quickly, for anyone.

I then help you with a novel approach to getting a “Customer 360 MVP” in production, quickly (MVP is a “lean” term for Minimum Viable Product). I then learn with you what directions make sense for your Customer 360 Journey. I am never “prescriptive” (do THIS, then THAT, using THIS solution). I am “proscriptive” (avoid doing THIS and THAT).

As I said above, the best place to start is solving the call center problem.

But, if you aren’t sure why to start I’ll show you where my previous client’s have started and how to manage a successful implementation. Read on.

Customer 360 Use Cases

• Do you need to access multiple different systems to get customer information while you are engaging them?
• Do you know what your customers are saying about you on social media, surveys, support emails and call records at the time of the interaction?
• Can you run analytics on all of the customer information you have to gain add’l insights on your customers. This drives both sides of the ledger…it saves money and generates new customer revenue streams. Some examples:
  • How much did you spend to acquire this customer?
  • How much is each incremental sale to this customer costing you?
• Would it be profitable to have a better, more comprehensive view of your customer?
• If your customer called your call center RIGHT NOW would your reps instantly know their customer history and be able to effectively upsell them?

What Customer 360 IS NOT

Vendors will try to sell you their product or service by calling it “Customer 360”. Be careful. Listen carefully, is this just some repackaged offering from 10 years ago?

A Single Customer View (SCV) aka A Customer System of Record

Some vendors want to take your customer data that probably resides in various CRMs, support systems, accounting databases, websites, etc and “integrate” it into whatever solution they are selling you. This seems similar to a data mart/warehouse. These are valuable tools to be sure. But these projects take months to years to implement and have a high risk of failure. Meanwhile, you haven’t solved your short-term pain.

Some people use “SCV” and “Customer 360” interchangeably. Traditionally SCV has been a “mastered” customer record that sits in one database and feeds all other systems that need customer data. That’s not Customer 360. As I just described it, SCV requires A LOT of ETL work. I believe that SCV is a panacea for sure, but probably unattainable. Instead, I focus on NOT copying/moving the data to the Customer 360 system, rather the Customer 360 should have real-time insights into the source systems. Read on for how I do this.

A CRM System

A CRM system is rarely a good choice for a C360 solution. But it is the first solution your CRM vendor will propose. Why? Because they want to sell you even more “vendor lock-in”. Don’t fall for it.

Your customer data probably primarily lives in your CRM anyway, right? But you probably also have valuable customer data in other systems…AR data in your accounting solution, browsing history in your website logs, even social media. How long will it take you to acquire that data and push it into your CRM? Weeks? Months?

I’ve seen clients spend inordinate amounts of time customizing and adding requisite fields to Salesforce objects like lead, contact, and account to try to leverage Salesforce as a C360 solution. I’ve never seen one of these projects actually complete.

And the data is only really valuable if it is near real-time. At the time the customer calls, you want to look at the shopping cart abandonment logs, not wait for the nightly batch feed into the CRM system. These solutions introduce “data movement” problems. Data Movement projects take forever and rarely succeed. This is no different than building another data mart. I recently wrote how you can avoid Data Movement Project Failures, those ideas work equally well with Customer 360 projects.

And I’ll bet that not every customer interaction occurs when your CRM system is in front of your user.

A Data Lake

I see “a data lake” as a proposed C360 solution a lot. Who proposes these solutions? Vendors with a vested interest in selling you more of their database or data lake product. What’s wrong with these solutions? The inference is again, “data movement”. It will take months, if not years, to copy your customer data from the source system to these data repositories. And will the data be real-time? And now you have to build the UI. And now you need data governance. It’s exhausting.

Master Data Management/Data Quality System

These solutions are proposed by MDM vendors. “Put your Customer Master Data into our system and we’ll manage your Customer 360 needs. We’ll even help you find Data Quality problems with your customer data. Finally, we’ll push the mastered data into every other system”. Do you see the implication? Data movement. How long will it take to build this solution? Will it be real-time?

That may sound harsh and your DBA will probably scream that the above assertion is WRONG, but it’s true. Marketers care more about getting timely data that may not be perfect over perfect data a week later. A really good C360 system can ultimately help you uncover REAL data quality issues, without obsessing over minutiae. Better data quality is A GOAL of Customer 360, but not THE GOAL, and we definitely don’t advocate using a MDM/DQS system to do it.

My Virtual Customer 360 Solution

Every Customer 360 solution will be different, but here’s the best way to start:

• a series of “screens” that can be embedded into any application that interacts with your customer.
• Provides lightning faster lookups to the customer data in the system where it LIVES (no data movement).
• Provides actionable insight at the moment of interaction (System 1 vs System 2 thinking)
• It is the source for Machine Learning systems
• this becomes foundational. It is the vehicle that helps you continue your Customer 360 journey.

These “screens” (I call them System 1 dashboards) can interface with underlying data from various source systems via REST APIs. The screens are actually REST calls themselves, which is much easier to embed in a Salesforce application than trying to move the underlying data to Salesforce and building the SOQL queries and interface there. With a “virtual Customer 360” solution you can now tackle your customer problems QUICKLY.

System 1 thinking is automatic and intuitive thought. It allows you to react quickly. This is exactly what you want for a Customer 360 dashboard. These dashboards are usually “red/yellow/green light” and filled with actionable information. System 2 thinking is more analytical and purposeful. It answers harder questions but requires more time to process. Marketers and call center reps prefer System 1 processes that cna give them something actionable at the time of the customer interaction.

Call Center Dashboard

This is the easiest place to start most Customer 360 journeys. What information would you actually display on your cs agent’s desktop if you could? The “screens” that you build can quickly display key information to your reps regardless of the CRM platform you use.

These systems avoid repetitive, costly work like re-keying of information. The customer is happy because the cs rep now knows the entire customer history, which helps to establish context. Pretty soon you are learning exactly HOW your reps are interacting with your customers and what methods work. You can now use ML to help your reps drive to desired outcomes.

“I’ve been tasked by my CMO to start a Customer 360 problem. Where do I start?”

I follow these best practices:

• Customer 360 projects fail when you try to “boil the ocean”. Start small by solving a simple use case: give your call center reps the information they need to make happy customers at the time of initial interaction.

“Build it, they will come”

This only works in Kevin Costner movies. Don’t force your users to use YOUR C360 system, integrate it into whatever they use today.

• Start small and experiment. You will have unique use cases. Create quick projects to learn what works. Don’t fear failure. Better to fail after 2 weeks than 2 years.
• Machine Learning is rarely a good initial candidate use case. It takes a lot of time to do data acquisition before you show value.

Next Steps in the Customer 360 Journey

Once you have a basic Customer 360 system (I advocate the System 1 Dashboard) you can quickly move to:

• Targeted marketing campaigns
  • Customer micro-segmentation
  • Campaign analytics
  • Geo-based analytics
• Churn and retention
  • What is the likelihood my customer will go to a competitor and what can I offer them so they won’t?
  • Price Elasticity effects on my customer’s propensity to buy
• Take Customer Service from Reactive to Proactive
  • An example: “Thanks for calling Mr. Smith. Are you calling about the product you just looked at on our website? Do you have any questions?”
  • Don’t wait for your customer to call you. Anticipate their complaints. Alert them real-time. Give your customer a dashboard.

Soup to Nuts: A Customer 360 Reference Implementation

Phase 1: Better Reactivity Phase

The goal of this phase is to react faster during the customer interaction. We prefer System 1 Dashboards for this.

• Avoid copying data. Instead, determine how to query data where it lives.
  • Avoid querying database data. Instead, try to query the system’s data objects. This avoids having to handcraft SQL queries.
  • Most systems have REST endpoints that can be queried by key and will return semantically-enriched objects with exactly the data you need.
• Create a single repository with the “customer data keys” to any system that has customer data we need.
  • Just store the “customer keys”, nothing else. A customer key is a customer id.
  • Every system has its own key and they probably aren’t the same between systems.
  • This repository “unifies” the keys, while leaving the important data where it lives today. Remember, avoid data movement
  • The repository needs to be lightning quick (in-memory, NoSQL databases work best) with a REST API front end that any data consumer can use, including your CRM system.
• The Customer 360 “application” becomes a series of “screens” that are used whenever and wherever there is a customer interaction.
  • At the time of customer interaction, the application simply makes a call to the C360 REST endpoint passing its “customer key”.
  • The C360 applicaton looks up that data in its NoSQL datastore then sends REST calls to the other systems
  • Those systems respond with their data and the C360 screen aggregates that data into useful dashboards the cs reps can use.

The above can usually be done in under a month. The screens are usually very basic, but this is good. This lets you experiment with look-and-feel and what the most important content is, quickly.

Start with acquiring the data that will matter the most: the RFM (Recency, Frequency, Monetary) data. This data will help you determine which customers are your best by focusing on the 3 factors that matter most. This is a quick effort.

Notice I haven’t told you EXACTLY HOW to implement this. Every client is different. During this phase you begin to see what the future looks like, you begin to see the overlapping and conflicting data issues that exist, you start to see why your customers are unhappy and why to fix it. What you haven’t done is built another data silo in another database that needs to be backed up. We haven’t implemented Hadoop or some monolithic data lake.

This is NOT an expensive undertaking.

Phase 2: Proactivity Phase

Phase 2 is different for every customer, but what I’ve found is that our clients want to make decisions and interactions faster at this stage. Here’s what I do:

• enrich the data in your in-memory repository. We store data that you determine is CRITICAL for a successful customer interaction.
• We use Machine Learning to pre-calculate what the cs rep should be saying to the customer if the customer interaction occurs

Phase 3: Predictive Phase

At this point all my clients have learned a significant amount about their customers and business. This phase, minimally, takes the data from the previous phases and makes it “predictive” from simply “reactive”.

• pre-calculate who you should contact NOW to make happy customers and increase revenue. Don’t wait for the customer to contact you.
• Data producer systems send events to a common data hub. Any system that is interested in that event, especially the Customer 360 system, can ingest it. I call this Kappa Architecture. See my lastest presentation.

Phase 4 and Beyond:

By this point I no longer have to make suggestions to clients. By now they have their own ideas and they are salivating to implement them. Some examples:

• Predict revenue and Customer Lifetime Value. Up-sell and cross-sell accordingly.
• Model how customers are sensitive to price elasticity. Amazon is really good at this. If you put something into your Amazon shopping cart but don’t buy it for a day, generally the price will be lowered.
• Other Machine Learning and AI opportunities unique to YOUR business

Why my Customer 360 Projects Don’t Fail

• I don’t focus on the hype. I don’t propose “data lakes” or “data warehouses” to solve THIS business problem. Not every customer needs one or has the staffing that can support it.
• Other vendors have “Know How”, I have “Know What” too. I listen to your problems and help you get started without proposing a expensive product.
• I don’t propose 18-24 month projects. I like to start small and ensure you are learning about your business. I follow Lean project cycles where we deliver small features and get your feedback.

Where does it hurt today?

What nagging customer problem do you have today? Or, think up a use case for customer data that isn’t currently possible for you today. Now think of how you could solve it with my approach. It doesn’t seem so daunting now, does it?

Up to table of contents