Team Bandwidth & Latency

Grow purposefully and uncover hidden value

I’m still playing around with content and format and this article might be a bit too jargony.  I’ll try to define terms before attaching them to the team concepts but I’m very open to feedback.  A lot of the topics I’m planning to cover in this blog are mental models I’ve built for myself as they relate to Startup growth, like any model it’s an approximation of reality.

Two people start transferring a massive download of 20 terabytes of data at the same time, one with their newly installed 1Gbps internet connection, the second by packaging up a 20TB harddrive and overnighting it to the destination.  Which data transfer mechanism has higher bandwidth?

The answer is the postal service, the internet connection will take a little over 50 Hours to complete.  Bandwidth is about total capacity over time as opposed to the speed of transfer.  The latency of both services is dramatically different, several milliseconds for the internet connection compared to a full day traveling by truck.  

A company of one person has the lowest latency it will ever have.  From there on out every new hire divides work and creates higher latency interactions with a small boost to the organization’s bandwidth.

Compete on your terms

When I joined ArchiveSocial we had a rough division between Sales & Product.  Everything that eventually became separate departments and processes fit across a few individuals with some post-it notes and spreadsheets.  In my first week I got a physical phone that routed all support calls to my desk.  Someone from a small town in Minnesota would call in, we’d have a quick chat about how I used to work a county over and they would tell me exactly what problem they were having.  With a few keystrokes we could usually get the problem fixed within the afternoon.  Low latency, low Bandwidth.

As a small company with a few customers you can out-compete larger organizations every day on latency.  This is extraordinarily valuable early on because those low latency interactions lead to faster iteration cycles and a product that is more fine tuned to the value being provided.  The language used matters here, it’s important to lean into these strengths in order to frame the conversations they will inevitably have with your larger competitors.

Since advertising is pervasive it’s easy to get caught in the trap of competing on the same terms as a company that is emphasizing their high bandwidth, high latency strengths.  Words like scaleable, reliable, trusted are all great words but they might come at the cost of a large team that has been working in a problem space for years.  Your early product might not be stable, but you can have a personal conversation and respond immediately to early customers and that will buy more good will than 24/7 support with a “high bandwidth” chatbot.

Scaling Customer Support

Being successful growing a customer base will inevitably require increasing the bandwidth of your team.  As a single developer on a support line, my ability to respond to emails and pick up the phone between coding led to a non-stop state of context shifts and impacted our customer experience. For us, we now knew common issues our clients had and we were able to offload some of that work by moving tasks to earlier stages of our pipeline.  We also solved it by hiring great people with the same passion for a personal human experience and growing a department focused on those customer interactions.

Growing the team’s bandwidth will always have a short term latency cost.  The instant access to information or deep product knowledge might require a Slack message or email.  The bugs that could be quickly addressed might need to be planned or prioritized.  By acknowledging this tradeoff we can start asking, “how do we keep latency for the customer low?”  This might mean creating a knowledge base or providing a direct communication path to people close to the product to raise new problems.

Value your team’s latency

As we grew I definitely made the mistake of holding onto our small company strengths far past when they exceeded my own bandwidth.  For me this led to a firefighting mentality where everything needed to be addressed with a low latency response, even if it wasn’t the thing that would provide the most value.  I give a lot of credit to the experienced people we brought in who knew that a simple bug backlog with regular prioritization meetings could solve some of these bandwidth issues while still prioritizing getting a quick low latency response to our users.

Unrecognized value

Customers or partners may have bought a product based on personalized support and low latency provided by a small company.  They experienced similar growing pains as the organization grew their processes and team.  This is super important feedback because it impacts what gets promised when you sell, how you tune those interactions, and ultimately whether to evolve the team in a way you can continue to provide that value.

In some cases I would rate the support being provided as a product in itself.  If you suddenly change the experience by increasing the latency the user will probably reevaluate their purchase leading to churn.  Imagine if Google search was suddenly 10 seconds, it is searching the whole internet.  You’re still receiving the same results you were before, but are you?

The latency trap

On the flip side, sometimes you might have a single customer (maybe a large one) who got sold a small company experience.  It’s important to not fall into a trap of limiting your overall market to meet these demands and not grow in a sustainable way.  Realizing this might mean a major shift for a business that has built a culture around these early principles.  

This type of change should be confronted as soon as you know it needs to happen and always starting with why.  Internally you might want to describe what not changing will look like 1 year out or 5 years out.  Externally having honest conversations with the early adopters who might have taken a chance on you or providing a clear roadmap of what the transition will look like.

In both places the people who signed up for a different mission or a different value proposition might look to exit.  In both cases having the conversations matter, a long and frustrating direction shift is less salvageable than selling people on a shifting mission.  This comes up a lot with acquisitions and mergers because different organizations will be used to a different bandwidth/latency tuning.

Purposeful Growth

I’m going to get into team growth and decoupling bandwidth from latency in a followup post.  I believe the two are linked but high performing organizations can separate out processes without sacrificing the important cross team collaborations that lead to better experiences.  Bandwidth and latency are two metrics which are highly related and tend to impact the other negatively.  These are the best types of metrics because tracking a change in both will ensure growth is not going to kill the parts of your system that might be hiding value.