Answered

How do you orchestrate your messaging?


Badge +3

How does the wider community approach the orchestration of interactions (email/post/chat)?

 

Challenges we have hit head on on include;

  • Targeting the correct users at a company
  • Limiting how many interactions
  • Landing the next best message
  • One Off Messages versus Ongoing Campaigns
  • Intercom versus other systems and not bombaring your customers
  • Visualising and mapping out the live touchpoints and interactions

 

We have come up with workarounds and ways to keep the lights on but not sure we really have solved this at scale - so keen on others experiences.

 

 

icon

Best answer by Anonymous 13 July 2020, 18:35

View original

6 replies

Userlevel 1

@Kate Sugrue​ I wonder if your marketing nous would be useful here?

 

@murat k​ could you offering Craig some insight here from your experience in using Intercom as a marketing tool?

@craig​ Hey! I will be happy to help as much as I can. But I need some specific scenarios about each. Could you give me some details?

Because the specific case will bring out the solution, I believe..

Hey @craig (SiteMinder)​ 👋 Like @murat k (Shopney Inc.)​ , the more info you share the more specific we can be with advice but I had a quick chat with one of Customer Engagement colleagues and we've put together some tips:

 

  1. Your targeting will only be as good as the data you have, so think of ways to get users to input their role/title, or whatever info you need, during the onboarding process. We've used the 'ask a question' app in the messenger to effectively gather this sort of data in the past, but you could also build it into your sign-up flow and use it to create a customer-data-attribute for targeting.
  2. Use ‘last contacted at’ for every campaign, message and bot. This filter is your best friend when attempting to prevent over-messaging. We also suggest that for campaigns, use ‘at least 3 days apart’ unless the campaign requires more immediate messaging, i.e. onboarding.
  3. To ensure users get the 'next best message' - create data points for key product actions, and define the sequence of actions. Trigger your messages based on the actions a user has or hasn’t taken. By defining the sequence, you’ll be able to target users based on the next action they should take to be successful.
  4. One-off vs ongoing - this depends on the whether the audience is static or rolling, and whether the message is evergreen or a once-off communication. For example, we use ongoing messages to onboard our users. These messages are evergreen, and the audience is rolling. If we had an outage that affected a certain group of users, we would send a once-off message to that static audience. For product announcements, we use ongoing messages within a campaign, as it allows us to reach more users as they qualify for the campaign, and to send a sequence of message to help educate users about different parts of the new product or feature over the course of weeks. (Top tip: users exit a campaign once they receive all the messages so we always add in a dummy message with rules that prevent it from sending (eg. email = fakeemail@fakeyfake), just in case we need to add further messaging to a campaign further down the line).
  5. Intercom vs other systems. We solely use Intercom for customer communications, but we do sometimes experience this issue on the prospect side where we use a number of systems. Our first piece of advice would be to develop internal governance. Define the role of each system, the types of leads or users they communicate to and at what stage. This will help your teams to understand what is used when, and why. It will also help you to see where the overlap occurs. The second tip, if possible, is to build data connections between your systems. It’s crucial that each system has similar info about a user, so that you don’t develop varying profiles for the same person, which can lead to an incoherent experience. Combining tip 1 & 2 should help you build an ecosystem where the tools communicate with, and complement, each other.
  6. When I began building out our customer onboarding campaigns a few years ago I actually drew them out on a scrap of paper 😂 It's probably easier to use a product like Lucidchart or Figma though.

 

It's worth mentioning that our team are working on some really exciting updates that will change a few of the above answers...but I'm not allowed say any more than that 🤫

Badge +3

Thanks @murat k​  - use cases generally;

  1. Customers during Onboarding whilst new features / updates rollout and normal business events happen (maintenance, outage etc)
  2. Customers in a beta/pilot and normal business events happen (product release, newsletters, maintenance, outage etc)
  3. Customers where data indicates they may buy / want more services, (product release, newsletters, maintenance, outage etc)

 

Targeting the correct users at a company

We have users that exist against multiple companies and that user is distinct as we use email as a primary key and do not allow for dupe users in our platform - as most engagement (sans ABM) is done at user level, we can and do send messages to multiple companies or to the wrong contextual company.

 

Limiting how many interactions and Landing the next best message

We also use last contacted and last seen to allow better filtering, but this may stop messaging being sent until that data point is not true (Eg. Last Contacted < 2 days) but does not help prioritize one over another.

 

  1. Critical Action Required
  2. Outage Impacting data
  3. General feature update
  4. General 3rd party partner update

 

If all message are live and waiting to fire, oldest goes first - but if we want to set the priority or order? What if the user is in a series, how do we prioritise messaging so the most important message goes first?

 

My use cases are more liml to running a SaaS business using Intercom as the engagement platform

Badge +3

I will go ahead and rate that best answer, with some further commentary inline;

 

1. The pain point is more our use case where someone could look after a patch of 50 hotels on any given day/week and log into and manage them but not solely be accountable or they are and they look after a group so i feel our data model of not allowing for many users for the same person has impacted us here

 

2. We do take this approach but with 5 products, many partners and many functions all needing to engage - messages queue up and we do not have a solve for that.

  • We have gone through using badges vs snippets vs posts and small or large to help educate users with deep page messagin
  • We also mix post vs chat for live vs passive
  • Could do with more types like notif bars, toasts etc

 

3. Not sure how you can codify actions or features as they would be contextual and linked to that point in the journey - unless everything had a coded / numbered sequence and you referred to A1 B7 C3 etc but the management of that on top of managing bots, series, tours etc gets out of hand.

 

4. Not sure what the **** message with rules means - but conceptually makes sense with what we want to do

 

5. We have the manual framework and clear ownership, the issue is feeding all that into systems so our teams know what is going on- so our gap is point 1 and 2 being connected - a 2 way SF/Intercom data sync in the new app would solve that 😇

 

6. Agreed, same approach here

 

Just on answer 4, it turns out our automatic moderator doesn't like the word 'd u m m y'. This is a word we use quite a bit on our team so @eric f11​ is going to remove it from the banned list 😊

Reply