The remarkable KPIs of mound builders
An introduction to semiosis and its relevance to organisational development
“A sign is something that stands for something to someone in some respect or capacity” — Charles Sanders Peirce
Macrotermes (large termites) are amazing creatures. They live in underground colonies throughout Africa and South East Asia. Inside dedicated chambers, they construct specialised gardens for cultivating fungus. They exploit the fungus to partially decompose their woody forage, making it easier for them to digest cellulose. Macrotermes are industrious mushroom farmers.
Macrotermes also construct large mounds out of soil. These towers, despite appearances, are not for housing. The termites live beneath them. The mounds are, in fact, an extended physiological ‘organ’. The combined metabolism of the termites and their fungi is equivalent to that of a large ungulate. Carbon dioxide needs to be exhausted from the colony, and oxygen replenished. Using an elaborate network of tunnels, the mounds harness the kinetic energy of prevailing winds to power the necessary gas exchange. The mounds act as ’collective lungs’ for the superorganism1. Macrotermes are also accomplished HVAC engineers.
Each colony and its mounds last for the duration of the queen’s lifespan, which is about 15 to 20 years. However, they are built, repaired and maintained through constant activity by the workers, who live only a few short months.
How do macrotermes perform such impressive feats of agriculture and engineering with such tiny brains and short lives? How do they coordinate their activities without a centralised bureaucracy? How do they perform complex work without architects, planners, blueprints and bosses?
In 1959, French entomologist Pierre-Paul Grassé introduced a beautiful term to describe the swarm intelligence witnessed in insect societies — stigmergy.
It is derived from the Greek words stigma ‘mark’ and ergon ‘work’. It captures the idea that agents leave ‘signs’ in the environment, and these signs subsequently direct the activity of other agents2.
A good example of stigmergy is how macrotermes repair their mounds. It is worth noting that the workers are blind.
When the outside of a mound is punctured, an abrupt change of atmospheric conditions is sensed by nearby termites. They immediately stop what they are doing and move about vigorously. Some race towards the source of the perturbation, while others, called tocsins (‘bell-ringers’), head in the opposite direction to recruit help. Soon, waves of termites, summoned from the depths of the colony, are swarming over the damage site. In the first stage of repair, the termites simply drop a dollop of soil randomly near the hole. The soil is laced with a cement pheromone. The pheromone is a sign. In the second phase, the termites, instead of dolloping soil randomly, deposit it where they detect pheromones. Pillars start to form. Some pillars develop stronger concentrations of pheromones than others. These are targeted with more deposits while pillars with weaker concentrations are abandoned. Thick pillars soon form which fuse together into walls and eventually become solid plugs.
The termites use their pheromone signs to focus construction activity on where it matters most. Their sign system enables them to filter out background noise and process only the most salient information. It is efficient and effective.
We use semiosis too. Our organisations are permeated with sign systems: metrics, KPIs, OKRs, analytics, dashboards, reports, meetings, etcetera4. These are supposed to help us focus on what is important, much like the macrotermes’ pheromones.
However, the sign systems belonging to most organisations fail to stimulate optimal behaviour. In a large study conducted by MIT Sloan Management Review and Google, 74% of executives didn’t think their functional KPIs were aligned with the organisation’s strategic objectives5:
most companies do not deploy KPIs rigorously for review or as drivers of change. In practice, KPIs are regarded as ‘key’ in name only; the most prevalent attitude toward them seems to be one of compliance, not commitment.
We are less adept at designing sign systems than macrotermes. Our problem stems from under-appreciating their role in collaborative work.
Signs are not decorations
We certainly desire good sign systems — there is massive industry feeding that desire — but we usually lack the intellectual discipline to pursue perfection.
Take this terrible advice from The KPI Institute, for example: ‘Selecting KPIs is like picking flowers from a field. You can't pick all of them. Instead, you have to decide on the right combination and a limited number to make a beautifully balanced bouquet... Using your nose generally helps.’ Perhaps the quote is taken out of context but it resonates with the lackadaisical approach to KPI design found in most organisations.
The basic principle of a semiotic process is that a sign informs an agent about the state of the world in which it lives, and the agent then uses that information to alter its behaviour6:
Sign → Interpretation → Behaviour
To design effective sign systems, it is best to start from the desired behaviour and work backwards.
Given a particular situation, what is the desired behaviour of the agents involved? Can the signals that agents receive be clearly interpreted? Do the agents know exactly what to do for each signal, or is there ambiguity? Is only salient information being transmitted, or is the channel full of noise?
These are the type of design questions that need to be addressed if stigmergic effects — spontaneous alignment — are to occur.
Organisations are based entirely on semiosis, on sign operations. If an organisation’s sign system becomes dysfunctional, then the organisation becomes dysfunctional. Dissatisfied customers will be left to fester. Disgruntled employees will leave. Threats will be downplayed. Opportunities ignored … The organisation will fail to adapt to its world.
On the other hand, by rigorously reworking and optimising their sign systems, organisations can start to unleash the full potential of their agents, enabling them to respond to threats and opportunities as autonomously as macrotermes.
For more about biosemiotics and semiosis in the natural world, the journal Biosemiotics is a terrific launchpad.
I will be discussing organisational semiotics in more depth in future posts.
Termite Mound: via bernswaelz, Pixabay
Macroterm mound diagram: Grassi B. e Sandias A. - Costituzione e sviluppo della società dei termitidi. Atti Accad. Gioenia Sc. Nat., Catania, 1893
Turner, J.S. Semiotics of a Superorganism. Biosemiotics 9, 85–102 (2016). https://doi.org/10.1007/s12304-016-9256-5
G. Theraulaz and E. Bonabeau, A Brief History of Stigmergy, in Artificial Life, vol. 5, no. 2, pp. 97-116, April 1999, doi: 10.1162/106454699568700.
Peirce’s semiotics is different from Saussure's semiology. See: https://en.wikipedia.org/wiki/Ferdinand_de_Saussure#Language_as_semiology
Semiosis also includes all of the non-explicit signs we exchange including our tone of voice, our body language, etc. For brevity, I’ll leave that discussion for future posts.
Leading with next-generation Key Performance indicators. Findings from the 2018 Strategic Measurement Global Executive Study and Research Project. MIT Sloane/ Google.