My friend Paulo Caroli wrote in his blog about a project that K21 carried out for a large client. The service provided was inspired by McKinsey’s horizons nomenclature.
McKinsey’s planning horizons:
H1: Short-term horizon. This is the dedication to the business that currently pays your bills.
H2: Medium-term horizon. Companies have to allocate part of their efforts into their next profitable endeavor.
H3: Long-term horizon. In this horizon, companies should invest in discovering what their future sources of income will be. This way they will be prepared for changes in the industry.
New Horizons Perspective
In practice, what we’ve realized is that, besides the issue of distributing one’s attention among these horizons, part of the effort is also invested in a past horizon. This happens every time we fix bugs or work on something that should have been ready. I refer to this as H0.
H0: Past horizon. Effort dedicated to something that should be ready, such as technical debt, fixes, manual tasks for lack of automation, etc.
Paulo especially liked the nomenclature we created when we correlated these horizons with our client’s service classes:
H0: DEVOLUTION
H1: EVOLUTION
H2: REVOLUTION
H3: INNOVATION
For example, for this client the following effort distribution was defined:
INNOVATION: A 10-person team dedicated to it
DEVOLUTION + EVOLUTION + REVOLUTION: 50 people (including managers) divided into 4 teams
For these 3 items, the following effort distribution was used:
40% devolution
40% evolution
20% revolution
Thus, there are 4 different backlogs. In some situations, it is indeed more beneficial to work with multiple backlogs – especially when there are different service classes. This helps in making decisions about prioritization, because we don’t have to compare all of the items when they have different natures.
It’s worth noting the strategic importance of discussing how the dedication to each of these classes is divided within the company. In the example above, the CTO was present when the percentages were decided.
Finally, the day-to-day control of capacity distribution is done using Kanban techniques. We also use Meddlers, a Management 3.0 tool, to decide how to allocate people on the teams.