Landmines! When your product abruptly fails your customers
Landmines! When your product abruptly fails your customers
Earlier I wrote about how moving upwards into the enterprise segment might reveal new gaps in your product and perhaps your entire GTM. There’s another phenomenon that can start rearing its ugly head during that phase: landmines. Landmines are these unexpected (and unpleasant) surprises that customers discover about your product. They are almost inevitably about misaligned expectations. The customer expects the product to do x, y and z and discovers that the product actually does a, b and c.
I describe this phenomenon as a landmine, because it is abrupt, unplanned and is painful for both the customer and the company. In this post I’ll cover when during the customer journey do these typically occur and what to do about them. I’ll also offer a few options to minimize the chances of them occurring.
When do landmines typically pop?
There are two phases in which these landmines can pop. The first shortly post-sales when the customer puts your product in production. Pre-production, your product might not have been exposed to all the workflows and use cases that the customer intends to use in a production environment. That’s a problem, which I will address in a later section. During this phase, if the customer discovers a gap between what your product does and how she intends to use it, that quickly escalates into a show-stopper. The recency of the purchase combined with the desire to start using the product and deriving value out of it from the customer’s perspective makes navigating landmines in this phase very challenging. 80% of the landmines I have experienced occur in this phase.
Another phase in which landmines can pop is when the customer changes, or introduces new workloads/use cases. These newly introduced use cases might not necessarily work well with your product and can end up causing another show-stopping escalation. These landmines aren’t all too common, maybe 10% of the ones I experienced. As always when they do, they are also challenging to navigate.
The last phase I’ve experienced landmines is when your product runs in degraded mode, usually in response to some failure. An example can better illustrate those.
Years ago one of my customers called me bitterly complaining about how her clustered file system behaved when a disk failed. Clustered file systems distribute data across more than one disk for redundancy (see Reed-Solomon). When a drive fails, the system starts recreating all the data that was persisted on this disk from other “copies” across many other disks. These file systems try to minimize the time to reproduce the data to reduce the data-loss risk window. That could come at the expense of the usability of the file system. That’s exactly what happened in this case. A disk failed, the system reacted and started using all available resources (networking + compute) to recreate the data but at the expense of clients (end users) being able to use the file system. From the customer’s perspective that constitutes an outage, but more worrying a behavior they cannot plan for.
“I cannot have hundreds of artists idle any time a disk fails. This must be fixed ASAP”
Boom! Landmine
What to do when landmines pop?
The short answer is when landmines pop, you almost always have to deal with them expediently and help out your customer. The only exception is if the ask - meaning what the customer expects from your product - is completely misaligned with your product. That implies that you’re better off not resolving this issue at the risk of losing the customer. Truthfully I have never encountered this option. Therefore, it isn’t about whether or not you will handle this severe escalation, but when: immediately or in the near future.
Your ability to negotiate a timely manner to deal with a landmine is both a function of its severity and the phase that it occurred in. A landmine in the earliest phase of product adoption is the hardest to negotiate around. You almost always have to deal with those immediately. Similarly, some of the landmines that can pop during degraded mode scenarios can be so severe and intrusive to your customer’s workflow that they will demand it to be resolved immediately. The ones that occur due to changes in workflow are the ones I have found to be easier to negotiate on time to resolve.
We’ve observed that most of these landmines require immediate treatment, which implies finding (engineering) resources to deal with them. That in turn disrupts what your software development teams are working on. You will have to stop some work-stream and pick up this escalation instead, which is hard, but necessary. Landmines aren’t fun.
If you are lucky, most of these escalations, although abrupt, can be short to deal with. There will be a class of them that are hard problems and might require months to resolve. If you find yourself in a situation where the customer requires an immediate or very short-term fix to a landmine, yet resolving it will require much longer than that, you need to start thinking of alternatives. These alternatives are almost always monetary and can come in extending the customer’s license or offering free services to recognize that the customer is unable to use your product until you resolve her issue. I once had to resort to leasing a competitor’s product and giving it to my customer for months to unblock her whilst my team dealt with a pretty severe landmine.
How to minimize the odds of landmines popping?
The onus is almost always on the company to minimize the chances of landmines from popping. In the case of the ones that occur during the early production phases, those can be mitigated by having a robust production validation or proof of concept stage before the customer makes her purchase. This is a process that should be led by your sales engineering team. They should ensure the technical validation of the product for the customer’s intended workloads and use cases. Landmines that pop during this stage signal a poor POC or technical validation process.
Similarly, landmines that pop when the customer changes her product use cases signal one of two cases: inadequate testing or customer education. If the newly introduced use case is one your product ought to support, then a landmine suggests improper testing on your part. On the other hand if the use case wasn’t what you intended your product to be used for, then this signals improper customer education. Either way, the onus is on the company and not the customer. It’s usually never the customer 🙂
The degraded mode landmines are 100% a function of poor testing. You need to test your product in various degraded mode scenarios and describe its behavior to set the right customer expectations. You should also consider the consequences of these behaviors on your customer’s workflows.
Things I am reading/listening to
AI for the Next Era - There’s a lot of hype around generative AI, but I think this hype is warranted. There could be many real, and useful, applications for this mode of AI. This podcast covers generative AI (DALL-E) and more generally the impact that very large models and AGI can have on our future.
This VP is Doing Things Differently in the Product Org — Here’s His Playbook - This is an interesting take on the role of the PM. It extends the role to own and be accountable for business outcomes like revenue targets.
I’ve recently started to read Ed’s work and have to say it’s fantastic.
