Measures and Signals as Taught By Bees 🐝

Happy Friday,

Yesterday I was out working with my bees for 2 hours. I'll have a YouTube video of it up later today, I hope. The reality is, I spent 2 hours out there because I got turned around when I went to do one thing, but wound up doing something else.

The work I did was splitting my 2 hives into 4. Originally, I was going out to feed and prepare to split next week. However, when I started to inspect to make sure my hives had enough resources to split, I saw that I needed to split immediately. While this isn't technically a big shift, I didn't plan for this and had to adjust my plans.

This brings me to my topic of this email: what beekeeping taught me about metrics and signals.

Let me first begin by listing the types of measures and metrics that would be "obvious" to a manager of bees or a consultant.

• Number of bees
• Queen age
• Number of eggs laid per day
• Frames of honey
• Pollen stored
• Number of supers (Boxes)
• Stings per visit
• Temperature
• Moisture
• Rate of food consumption
• Cost of ownership
• Time to inspect
• Time to light smoker
• Mite count
• Drone %

I can keep going, but you get the idea. I could even ask for estimates of these too. Then, when the estimates don't work, I can spend time judging my bees for not meeting the projected estimates.

Sounds silly, I know.

Now, the real question is: Is there merit to any of these measures and metrics? The answer is: Yes, depending on the situation.

So now I'll reframe things in terms of the actions or decisions I need to take as a beekeeper:

• Do I need to requeen my hive?
• Do I need to split my hive?
• Do I need to treat my hive for pests or disease?
• Do my bees need food?
• Do I think they'll survive the winter?
• Do I need to combine hives?
• Should I harvest honey now?
• Do I need to give them more room?

There is more of course, but this is enough for me to get to my point. If I look at these actions or decisions, now some of that data above can help. It is the connection of the decision to the number that matters.

Let me dive into one of the most common things I do as a beekeeper: Check if the hive is "queen-right". That just means the hive has a functioning queen. Without one, the hive will die. It takes 16 days for the hive to breed a new queen, and she has to then fly away to mate, which carries a chance she dies, and that may only work at certain times of year.

That data feels less and less useful given all these variables, doesn't it?

So, here's what I actually do as a beekeeper. I look for signals that I have a healthy laying queen. One signal is to see all stages of life from egg to capped brood cells. This provides an indication that I have had a queen laying eggs from 21 days until 1-2 days ago. The next signal is to see the queen.

Is this perfect? No. I can see a queen, but she can be new and unmated and die the next day. I can see eggs, but the queen died 1-2 days earlier.

So what do I do? I will inspect again in 7-10 days. Now it will be obvious between these two inspections as to the situation.

This is the most common thing you inspect hives for, and we use signals to give us an indication as to what we need to do.

I don't spend time looking at a spreadsheet of data where numbers fluctuate, and I then try to divine meaning from it.

My job splitting hives today is an example of this. I inspected for queen-right, and the overall set of resources each hive had to indicate that I could safely split. The surprise was seeing queen cells that will hatch. They had the cells, the resources, and active queens. Time to split.

I want my splits to have 2-3 frames of brood and 2-3 frames of food. Simple to count and assess in the moment. It is a signal, nothing more. If my hives have those resources to spare, I can split. If they don't, can I borrow from a stronger hive?

One of the bigger mistakes I've seen in my clients as it relates to data is that they never develop an explicit signal-to-data relationship. They are looking at data and trying to create meaning from it. The opposite is what works better. Know what actions you need to take, then what data will indicate that you should act.

If got data and dashboards but struggling with insight, reply back and let me know.

Sincerely,

Ryan

PS: I'm hosting a free webinar on May 21st called "Lies our Engineering Metrics Tell Us" I'd love for you to join and spread the word.

PPS: Also, I'm planning to offer private workshop for engineering leadership teams to help them more effectively use data to inform their decisions. Rumor has it, the illustrious Esther Derby will be joining me!