What are your key takeaways/must-do's when implementing a CS software?

I agree with the previous comments. I do want to echo one of them: you do need to think through your processes and how it works with the system you're implementing. Previously, we implemented a CS system without scrutinizing our own processes and how the system would work with them. During our original implementation, we didn't do this very well and mostly redirected our existing manual processes into a separate system. Fast forward a few years later, we're revamping most of those efforts. In addition to time lost, people got used to the old way of doing things and those legacy practices were calcified simply due to the way we had our systems set up. 

In addition just as you would when managing change with a customer, you want to create quick wins for your team. Without it, most CSMs on the team will see little value in it. As part of our revamp, we have been able to make quick wins: dashboards to help review accounts, automated data that saves time/effort, alerts, etc. This helps with buy in and general positive sentiment as well as when you're looking for feedback and input from the team.

Thank you @Ed Powers @Fleur Duekker @Chad Horenfeldt @Dave Epperly @Srikrishnan Ganesan  for all your thoughts!!

I agree with Chad's approach. I would also pick a single use case that is the biggest problem you are trying to solve. I would recommend solving it from a CSM angle first before a leadership angle to create buy-in. Once you get the biggest problem tackled then start to layer in the other use cases.

Hi Peter would love to help but in infancy in my organisation. Currently moving to a unified CRM which hasn't been established so really early days but with a PM background 'discovery' in terms of the 4P's was first, significant consultation to try to bring the organisation along simultaneously, mapping current state and nirvana first and allowing the planning time. Now working with research to understand how that might inform the solution. Will watch with interest in replies as keen to understand what tool you go with as well.

Hey @Peter Sterkenburg--

To quote an adage, "If you have a mess and you automate it, what you get is an automated mess." This has been true throughout my career, regardless of the process or platform. The CS deployments I've seen are no different. 

I've found it's best to fully document and practice your process on paper first, including having metrics and controls in place, before doing anything. As @Chad Horenfeldt points out below, the benefits of automation then become very clear because you have established a baseline from which to measure and justify investments for improvement (better dashboards, higher CSM productivity, etc.) that come from automation.

-Write out the high level objectives and create a narrative on why you need a CS platform. This should be based on data: churn data, CSAT, VOC and NPS data, team engagement data. Consider where your company is going. How will this allow you to scale, reduce costs etc...
-Start to list out the use cases as to why you need the platform. Use cases are critical. They will guide you on what vendor makes most sense. 
-Get ops help. This is a must for larger CS teams and more complex tools
-List out the metrics you want and get your prod/eng teams aligned around how you will get this in the CS platform. Don't pass go before you do this. Cross-functional alignment is critical.
-Get your team involved. Focus on their strengths and interests. eg: getting help creating the health score. Don't take this all on yourself. 
-Create a team for the implementation that will help you roll it out. It's best to get some people on your team involved in the testing phase so they can help you with the change management. They can convince their peers to use the new tool

Experiencing bullet 1 first hand. Can confirm this is a real issue.

Hey Peter,

I've seen one CS implementation - but from the sidelines. I can share a couple of areas I've seen as problematic.  
* While it was easier to connect some of our peripheral systems (like CRM, billing system data, etc), it was hard to get the product team aligned to prioritise sending the right usage events to the CS tool. If you have something like Heap + Segment, might be better for you to send events to the CS tool than depending on product engineers to do this. This readiness and bandwidth was the greatest blocker for unlocking value. 
* Don't go in with trying to get everything working on day 1 - that's hard and will push your go-live out further. Instead define what could be a reasonable phase 1 go-live. Identify low hanging fruit in terms of health scorecards, alerts, notifications, automations, etc and get them done as step 1. You can do a workshop with your CS vendor to get this done in just 3-4 weeks from what I've read / seen.