Communications Plan: Nuances, Pitfalls, Solutions

How to automate and personify a marketing strategy with the help of a communications plan

Original illustration by A. Malyavina

A communications plan is developed by reaching a compromise. We strive to do the job well using standard strategies and our own expertise, but we also have to adjust them to the needs of a particular business. As a rule, these needs clip our wings and try our patience (“Couldn’t you just synchronize the cart both on the website and in the app?!”), but that’s okay. It’s not okay to act like all online stores are the same, or all real estate developers are the same, etc. It’s not okay to shrink away from studying the business and assume a spherical cow in a vacuum.

Your communication plan should consider the details of a tech ecosystem, business processes, and the internal policy of a project. It should make it possible for different people working with practical tasks to understand the physical significance implied.

Solving practical tasks

Let’s take a standard strategy — the abandoned cart. Communication plans usually illustrate it in the following way:

Abandoned cart

Now we’ll see how this scheme can be interpreted depending on the task.

An analyst looks for a window of opportunity for optimization

But this scheme doesn’t help with even the plainest of options:

• Should we give a coupon in the second email? Is there any now?
• Should we change the product recommendation? Which one is there now?

A technical expert starts the implementation process

And they don’t have enough data for even the basic settings provided by the platform:

• Should the trigger email be sent every time the cart session ends? Which limitations should be added, if any?
• The platform has two lists of carts: on the website and in the app. Which one is meant here?

These issues are not only technical — these are issues pertaining to business as well. So the superficial view given by the scheme above is not enough for a discussion with a business leader. Also, if your plan lacks the details, it’s hard to estimate the cost of implementing it.

It would be nice if clients kept the scheme updated after the implementation. Communications plans not only give a concept of how everything works but also provide a basis for planned changes.

If there isn’t enough data to solve practical tasks — the plan proves unequal to the task.

Design of a plan

The plan should include everything you consider significant for the project, for example:

• The maximum send frequency of a trigger email per user;
• The sender (e.g., CEO, the team, a personal manager);
• If the communication is compatible with different user segments (by last purchase data, by engagement in the channel);
• Whether the model of communication is marketing or transactional;
• Coupons (if any);
• Product recommendation and a basic section of products for it (“personal recommendations for the section of boys’ t-shirts”);
• Unobvious dynamic parameters (storage time of the order, tracking number);
• UTM tags (at least at the level of the campaign);
• The links to these communications on the ESP / CDP (if implemented);
• Priority settings of the email workflows.

Try to make the plan as unambiguous as possible.

Explain your concept of a certain email

If you generally use gender and age segmentation for the project, but a certain email lacks these settings — confirm that you’ve deliberately made this change.

Confirm the changes in segmentation

If it’s possible to use the same email for different workflows, you should mention this.

Mention that you use the same email for various workflows

State the starting point of time delays

Once I came across a funnel with communications on two platforms: AMO CRM and Unisender. Time delays were stated in the scheme with the use of AMO terms: from the moment of a trigger event. The implementation went okay in AMO. However, the funnel wasn’t fit for Unisender (in Unisender, the workflow design wizard time delays are not counted off the initial event but off the previous one), so +30 days from the status update in the funnel turned into +30 days from the previous email of the workflow, and the whole workflow lasted for 75 days instead of 30.

You can avoid such situations if you add a short explanation to your plan:

Specify the time intervals in the comments

Let the plan demonstrate manual communications as well as those implemented by third-party services

For instance, customer service for contractors presupposes that in case a meeting is canceled by the manager, they should write a personal message to the client. The plan should demonstrate this communication using a certain note to make it clear.

One more example: a notification of the order delivery to the pickup location is sent by delivery companies — that’s enough for an online shop. The plan should include such notifications to clarify that you haven’t forgotten about them.

Each project has special features which determine how detailed and complex the communication plan will be.

Of course, each project is unique, and each time you decide whether to add something to the plan, discuss it, do it manually, demonstrate it in Google Docs, etc. But this issue should be managed and agreed upon.

The physical significance of the plan: what you should pay attention to while implementing it

If your communication plan looks as if someone erased the details from it, it’s not always about the design. It’s okay if the author hasn’t added the details. It’s much worse if they haven’t worked them out.

Dig into the project, and the physical significance of the plan will be formulated on its own — the only thing left will be to design it.

When should you design the plan?

This question makes sense in the case of a project launch or a change in the platform. In such situations, the best time is when you’ve already chosen the ESP/CDP, but haven’t created the integration yet.

If the platform has been selected

We can make a scheme of communication processes for implementation based on special features of a particular platform:

1. Take into consideration the algorithms of the platform. For instance, Salesforce has a feature, “Einstein Best Action.” It allows to guide the users to the most optimal course automatically.
2. Let the plan demonstrate standard settings and those which are mandatory for this platform. For example, the frequency of triggers sent per user:
3. Consider the abilities of the platform: you can jump at the idea of sending special offers to those who shop on weekends, or abandon it and feel free to integrate something new.

It will be best if the integration takes place after your plan is designed

Our communication plan will be the basis helping us transfer all of the necessary characteristics during the first iteration of the platform implementation process. Remember: integration update won’t happen any time soon or likely ever.

If you start designing your plan at the stage when you have your integration ready, specify what characteristics and statuses you have. This will prevent you from building castles in the air.

Study the past, stay updated on the plans

Find out whether the project has a loyalty program, an app, offline shops, B2B and B2C marketing, saved items lists, and gift voucher feature (if not, whether they are going to be implemented in the future). Also, study how these features are integrated with each other and the platform: whether the loyalty program is valid in offline stores; whether the cart on the website is in sync with that in the app; if offline orders are transferred to the platform, etc.

Of course, it’s better to find out all of this information before you start doing the job: since every task should be included in the plan, it’s important to be fully prepared for it.

Request user survey reports that were carried out before you joined the project. They might help you start thinking in the right way: e.g., how to work out the customer’s rejection in the first purchase trigger chain.

Use the data, but remember the general logic of the business. RFM analysis can demonstrate whether order gaps are okay or when it’s time to start worrying about them. However, in case the reward points expire in a year and you need to determine the size of an order gap enough to consider a customer lost — don’t dig into the data. Let it be a year.

Find out whether the project has a global control group feature

If it does, a part of the email list goes to this group and gets excluded from the newsletter. Transactional emails ignore the global control group. But remember that there are non-transactional communications that should also ignore the global control group. If you show the users a pop-up with a coupon, you promise them to send it. It would be wrong if a user didn’t receive it because they’re in the control group.

Look at this scheme: the first email ignores the global control group, and the second one doesn’t. This is shown with a marker and explained in the comments section.

The first email ignores the global control group, and the second one doesn’t

Figure out the physical significance of the funnel statuses

You need to understand the following:

• What does each status mean according to the developers of the funnel?
• How do the managers understand and use it?
• How often does your platform get updated?

Once I came across the following case: a status was labeled “meeting”; when it was assigned, the user received an email notification with a date and time of the meeting. However, in the CRM this status was assigned after the meeting took place.

Automate business processes instead of creating them

Let’s imagine your client teaches online classes. They ask you to design a plan to automate communications. If you get excited and start adding notifications “You have a class tomorrow at 3 p.m., here’s the link” — just stop. You should learn how the process is managed first and see the funnel in the CRM. There’s a chance that the process is not managed at all and there’s no CRM; the teacher creates a link to a Zoom meeting three minutes before the class, restarts it in 40 minutes, and sends all this to the chat (not even to the children, but to their parents):

A screenshot of a thread made by the author

You see, you have nothing to automate here. Of course, you can work out and describe a business process, develop CRM strategies, write a guide for the staff, manage their motivation system, and set up monitoring… But this will be a slightly different task, including both the development and implementation of a business process.

When you start creating a communication plan, the first stage should be the analysis of the current business and search for communications you can automate.

Learn the business policy

Sometimes multi-brand projects require the removal of internal rivalry between the brands. It’s especially true for real estate developers: if the customer is in the funnel of one block, they shouldn’t receive emails about the other one. This approach requires additional settings in the triggers and product recommendations.

The same goes for the question of whether the newsletters encourage visiting offline stores or an online store only.

One more example — segmentation with different parameters (for instance, gender segmentation can be necessary for clothes, but unacceptable for toys).

Use segmentation for 100% of the users

We all love segmentation, but not everyone understands that some users can have irrelevant or simply unspecified criteria. Keep these users in mind.

When planning communications for the residents, think over how the filter should be set up: “city = New York City” or “city unknown OR New York City”. Or you can make a separate communication for the users with an unspecified city (taking into consideration the possibility that this city can turn out to be either New York City or something else).

Work towards a long-term effect

The sequence principles may be transferred from the scheme to the platform, UTM tags, and statistics dashboards.

Communications can be classified in different ways. When organizing the abandoned cart and saved items list, you can rely on the following aspects:

• The user’s actions (a session with an abandoned cart, but without placing an order);
• Price drop events;
• Shopping Cart;
• Wishlist;
• Subscribers lists.

Organize the abandoned cart and saved items list

There’s no correct way to classify the users’ actions. The best is the one which allows you to understand and analyze the statistics.

Don’t multiply the entities

Imagine you’re sending an email offering to fill in the account information as a part of a welcome chain and the same email after the order placement. If you need to distinguish stats between these two, create two triggers and two emails. If you don’t, then you’ll need two triggers and one email.

Let’s assume our project has two entry points (sign-up on the website and a subscription in the pop-up). Aiming at customer journey personalization, you decide to make two kinds of welcome sequences, based on the lead source. But the project keeps growing: in a month there will be three more pop-ups, and in a year the subscription will appear in offline stores. How many variations of your welcome campaigns will you have to do then?

The answer is… only one. You should make one sequence, based on a subscription status change. If offline clients were promised a gift coupon and the pop-up offered a discount, you can send additional emails after the basic series.

Look for universal solutions. Every crutch reduces the chances of the scheme to work effectively in the long run. You can’t improve the plan if it’s based on entity instances, not on entities — it just should be redone.

P.S.

There’s no universal standard for communication plans. Let’s work on it!