Outbound Decisioning
Fueling outbound marketing with an NBA strategy
Traditional outbound marketing defines a marketing action and then targets a group of customers for that action. More modern techniques basically do the same, but have more sophisticated ways of identifying who to target and maybe even at what moment to perform that action. That process is the exact opposite of having a customer first mindset. These campaigns can still generate a lot of value for a business when done well. That said, they shouldn’t be your default approach to reaching your customers in a context rich environment. If your goal is to optimize at a 1-to-1 level and generate the absolute most value you can with your marketing activity, you need to change your approach to do those things. Save those well researched and creatively orchestrated marketing campaigns for scenarios where you don’t have enough information to do better. When dealing with your existing customer base or known prospects, leverage all the information and individual context you have to optimize specifically for that individual at that moment. A fantastic way to do that is by integrating your outbound marketing efforts with an NBA strategy. If you aren’t familiar with decisioning platforms and next best action (NBA), skip to the bottom of the page and check out the introductory posts I linked on those subjects. Assuming you’re interested in optimizing your marketing interactions, I think they’re worth your time.
Basic outbound concepts in general can seem conflicting to an NBA based decisioning strategy. Fundamentally the mindsets between the two are different, at least for the majority of outbound philosophies. When a customer comes to us through any inbound channel we are given a clear point where we must make a decision. The customer decides the time and place and we decide how to react. The nature of those inbound interactions make it easy to think about the customer first, and to do so on a 1-to-1 level. Outbound doesn’t present us with the same simplicity. We have to decide it all. That’s why it leads us to think action first and customer second. Even when managing sophisticated targeting towards small customer groups, outbound actions are evaluated in a relative vacuum. We typically prescribe those actions based on generalized behavioral patterns, including when developing complex journey based campaigns or content streams. Proper use of analytics and machine learning can do a great job of helping us only pick the best possible targets for those marketing campaigns. They can even help us analyze target quality with at a more 1-to-1 level. Really complex strategies may even select between a few variants of the targeted marketing action to make some optimizations or personalization. Even those sophisticated campaign strategies typically fail to evaluate the prescribed action(s) in the full context of alternative actions. To do that a campaign needs to consider all the outbound actions in our ecosystem for each individual person at the given point. This gap in traditional outbound techniques leads to holes in optimizing at a 1-to-1 level for each and every interaction.
The hurdle of flipping to a fully customer/individual first mindset can make it difficult to design outbound marketing strategies that leverage NBA evaluation. Even companies with robust and sophisticated NBA strategies on their inbound channels can struggle to integrate outbound touches. Tackling that initial hurdle does come with its rewards. It will create the opportunity to eliminate that gap in 1-to-1 optimization. It should also ultimately result in a wholistic outbound approach that is much simpler from an operational standpoint to manage. Allowing you to consolidate a web of different campaigns with competing strategies into far fewer NBA based campaigns. Below we’ll walk through a few basic levels of organizing your outbound strategies within an NBA framework for optimization. You may have differing amounts of flexibility for actually managing outbound marketing within your specific decisioning platform. Regardless, applying an NBA based decision process can be valuable for any organization’s outbound marketing.
Quick disclaimer here: Modern marketing means a lot of different systems managing different pieces of the puzzle needed to execute a campaign. Everything from managing creative assets to executing messaging. Decisioning platforms are an excellent solution for bringing all of those pieces together under a single cohesive strategy and view of the customer. We’re not going to cover the technology and platform integration considerations here, but instead stick to the strategic marketing and campaign design elements.
Now let’s look at a few points across the spectrum for merging your outbound targeting with an NBA decisioning strategy. This NBA strategy could be a new NBA strategy, but if you have an existing NBA process designed for inbound channels, use it. This keeps your optimization goals consistent. Some updates can be made to consider the new outbound context, but the overall process should be consistent. Especially your scoring process. This ensures your optimizing towards the same goal.
No NBA Integration
There’s obviously not a ton to discuss here, but I think it’s important to set our baseline before we move forward. As the header indicates, at one extreme is no integration at all. Operating your outbound marketing in this manner doesn’t mean that you can’t orchestrate those interactions within a decisioning platform. It means that an NBA process isn’t used to decide who gets what touches and when. Some platforms are quite capable of managing complex outbound marketing campaigns in a way that allows you to do all your complex targeting and scheduling in one centralized location efficiently. Using those platforms in that manner can still add a lot of value to your marketing operation. More importantly it opens the door to moving towards 1-to-1 optimization for those outbound campaigns in the future. Don’t hesitate to start here if you need your decisioning platform to begin orchestrating some of your existing marketing strategies before you can consider reimagining them.
Partial Integration
Partial Integration is that middle ground across the spectrum of integrating your outbound activities with your NBA strategy. In this middle ground the decision to act in a general way is directed by the business for a customer. These customers are usually targeted because they are part of a broader group of customers that fit general criteria outlining customers the business wants to target with this type of outbound touch. Those targeting rules may be based on who it has been successful with in the past. They may also be designed to identify customers that fit a customer profile piece of context the group of actions were designed to respond to. We’ll refer to these general directives as campaigns. Similar to a typical marketing campaign just designed a bit less restrictively than a campaign with zero NBA integration. These partially integrated campaigns don’t prescribe an exact action isn’t by group level targeting rules. Instead they optimize between a pool of actions relevant to the campaign’s intent using your NBA strategy. Those actions can be all relatively similar, for example small offer variants. They could also be wide ranging. The less restrictive that campaign’s intent and the less restrictive your pool of actions the more you allow for optimization at a 1-to-1 level. It also moves your optimization in the direction of the consistent overall business goal evaluated by the NBA scoring process. To better understand what a partially integrated campaign may look like, let’s look at a simple example. We’ll go back to the hypothetical consumer bank from previous posts to do so.
DBM Bank runs a daily campaign to target their existing customer base for new credit cards. This campaign starts with a few basic targeting rules to identify customers who are eligible for the outbound campaign. Customers who aren’t in a high level arrears on existing cards, haven’t opted out of promotional messaging, and who haven’t received a credit card offer in the last 2 weeks through any other channel are selected. From there are all possible credit card offers are evaluated by the NBA strategy. We’ll assume a filter -> score -> rank & select NBA framework. The credit card promotions the customer is ineligible for are filtered out, the eligible offers are scored according to the prioritization formula, the optimal credit card offer for the customer is selected and sent to the customer.
This example is a pretty straight forward middle ground between completely orchestrating an outbound campaign and letting your NBA framework optimize your outbound activity. The core of the action is decided by a fairly specific marketing strategy. Within the confines of a range of credit card offers at a specific time the NBA framework optimizes the specific action to the one that best matches the overall optimization target. The more the bank frees up those restrictions the more individually optimized these customer interactions can be and the higher the level of integration with the NBA strategy. For example, a general cross-sell campaign that evaluates credit card offers as well as other credit products the customer is eligible for. Simple campaigns like this are a great way to get comfortable giving more of the decision making power to your optimization process.
Full Integration
The last level of merging outbound touches we’re going to discuss requires a larger shift in how you design outbound efforts. For the purpose of this discussion we’ll call it full integration, but there is a lot of room in between the type of strategy we’ll we’ll discuss here and our last example. There will always be perfectly reasonable use cases for campaigns that leverage your NBA strategy in a more restricted manner. We could talk ourselves in circles about how those use cases could, or even should, be handled with a pure NBA strategy. Some juice just isn’t worth the squeeze in a real world application.
For a fully integrated campaign nothing is orchestrated with any targeting or segmentation rules upfront. The basic strategy for this type of outbound campaign is simple because it leaves the targeting complexity to your NBA strategy. Like we discussed above, NBA based outbound marketing is centered on a customer first mindset and identifying the most optimal action to take for that customer. This is also occurring within the context of all the other possible actions that could be taken to select the most optimal action to perform. Done right this can allow you to organize a vast array of potential marketing actions without managing a complex web of competing priorities. We’ll spend quite a bit more time on this type of approach in order to explore a few design considerations in more depth.
Using your NBA strategy for making the action decision means the typical targeting and optimization decisions are taken care of at a 1-to-1 level, just like your inbound channels.
- The filtering process eliminates offers that are illogical to apply to the situation, for example offering pre-approval and low intro APR for a credit card they already have.
- The scoring process is used to evaluate all the remaining possible actions for each individual customer against your optimization goal.
- The ranking & selection process compares those potential actions according their score and selects the most optimal action(s).
The extra complexity for outbound marketing strategy comes from identifying points to make those decisions and create interactions. These decision points can come in a variety of ways. Two simple approaches for identifying them are triggers and time based scheduling. Triggers can be evaluated in real time or can be delayed and batched. In either case you are identifying a behavior (typically from the individual) and using it as reason to assess how to react to the new context with an outbound touch. Time based scheduling is as simple as it sounds, it’s using calendar based timings as decision points to evaluate customer groups with your NBA process.
Like any other outbound strategy, touches have to be balanced by a contact frequency policy. This way you don’t flood your customers with marketing touches to the point of negative impact. However, there can be a higher demand to understand those policies and their impact when scheduling a frequently evaluated outbound NBA strategy. Let’s consider the following example of a simple campaign in relation to the below outlined contact policy. This campaign runs all customers through an NBA strategy and is evaluated daily. All possible actions in your ecosystem are available for the NBA strategy to consider.
Contact Policy Rules:
- Limit optional outbound marketing messages to 1 every 3 days.
- Limit the number of times a single optional action can be taken to 1 time every 30 days.
It’s easy to see in an example like this how you could quickly fall into a repeated pattern for an individual customer. Every 3rd day a customer would get a marketing action, assuming nothing has changed that would result in them no longer qualifying for any of the action in the NBA strategy. Depending on the size of your action pool and how sensitive your scoring is to small context changes, it’s even possible to fall into a pattern of cycling through the same 10 or so actions in a repeatable 30 day cycle for a specific customer. Having some pattern potential is unavoidable when working on a schedule for outbound, and not necessarily bad. This example shows what can happen when you don’t consider scheduling and contact policies together. It’s not a problem unique to NBA campaigns, any marketing environment with numerous regularly occuring campaigns can encounter it. It can certainly be managed better than our simple example by using a more robust contact policy. An NBA strategy based campaign does offer the ability to mitigate these things at a more one to one level. By having a scoring process that is cognizant of the context from recent actions and customer reactions it can account for them. This allows you to take some of the general optimization of contact rules and move those evaluations to a 1-to-1 level. Including propensity models that factor in that recent context will mean the score for actions in a repetitive environment can be decreased at varying rates for each specific interaction. This alone won’t eliminate needing contact restrictions to act as a general governor, but it will help balance the mix of marketing actions to an individual. A useful tool for managing overall contact frequency at a 1-to-1 level is to consider the negative impact potential for taking an action in your scoring process. When the expected negative impact outweighs the expected positive impact for all actions your best actions becomes no action. This does add some complexity to your scoring process, but don’t rule it out as a strategy to combat our tendency as marketers to be far too optimistic. We often forget that there is risk in interrupting the status quo.
For those of you who already have a decisioning platform and NBA strategy, I hope you will consider leveraging those unique capabilities to elevate your outbound marketing. There is no one size fits all solution for integrating NBA decisioning into your outbound marketing strategy. Most marketing organizations will have use cases for all the levels of integration we discussed. Especially in the circumstances of low context, for example new customer prospecting, little to no integration with your NBA strategy can be the best option. It can also be a viable strategy for first managing your existing marketing campaigns as they were designed with your new decisioning platform. However, if your goal is to optimize all of your customer interactions at a1-to-1 level in all your channels, your outbound strategy needs to evolve. Do that by moving more and more of those outbound marketing touches along the spectrum towards a fuller integration with your NBA framework.
If your interested in digging more into how to merge various outbound marketing philosophies with decisioning and next best action selection follow me here and connect on LinkedIn. I’ll be discussing that topic at various points throughout the year, starting with journey based marketing orchestration. We’ll also continue to dig into techniques for the scoring portion of an NBA framework and designing those inputs.
Need an introduction to decisioning and next best action? Try this!
