Putting all your eggs in one basket

A common English phrase is:

Don’t put all your eggs in one basket

English proverb

This is sensible advice and is really suggesting a risk mitigation strategy (see Risk Management). The proverb, fairly obviously, means don’t put all your resources (money, time, energy) into the same activity, just in case that activity fails. Or, put another way, avoid single points of failure. Or, to use another common phrase, keeping your options open.

This advice applies in many situations and at many levels. For example:

  • Royal succession
  • Personal investments
  • Infrastructure
  • Portfolio management
  • Programme or Project contingency
  • Technical design

The Royals apply this advice in their efforts to have an "heir and a spare" to ensure succession.
At a more mundane level a sensible share portfolio has diverse investments to reduce the risk of failure of any one.

As my main interest is programmes, portfolios and projects I’ll concentrate on the last examples.

Infrastructure

To achieve resilience in technical infrastructure you must eliminate single points of failure. This applies when talking about servers, hence clustering, and databases, hence database failover. Even backups are a mitigation strategy in this mould; if the database fails you need a backup to ensure you can get up and running again.

Portfolio Management

Product portfolio managers have a similar succession problem to the Royals. If the main product fails, usually that means doesn’t generate revenue, they need to find another product to fill the gap. Intel had to face this when their memory business began to lose out to competitors (Grove, 1999). Luckily for Intel they had already started to invest in the emerging area of microprocessors. When management realised that the weren’t going to succeed in the memory business they switched tack and went on to become market leaders in microprocessor production.

Programme and Project Contingency

In the world of programme and project managers having more than “one basket” means having a contingency plan. For example the investment case for my current programme assumed the programme would use expensive contractors for the first year but then transition to cheaper, more-or-less, permanent staff for the second year. Given the skills involved, MS Sharepoint, were in high demand and carried a high price tag the investment case acknowledged that this transition might not be possible so the budget included an explicit contingency pot to cover retaining the contractors. As it turned out the transition was only partially possible so I had to dip into that particular contingency fund.

Technical Designs

I have also used this approach to risk management when tackling uncertain technical designs. This really just means exploring more than one design option. For example, on one programme we had get the subtitles of a video through the production system as metadata alongside the video content itself so that we could index the subtitles for search. Because of the complexity of the production chain, and the fact my team didn’t control the entire chain, we explored two options to solve this problem. The better solution was to capture the subtitles as metadata at source and route them through the production system. This involved a piece of bespoke work by a 3rd party software team and work by two software teams within the organisation but not part of my programme. With dependencies on three software teams, all outside my control, this looked high risk to me so we explored another, less elegant, solution. The side bet relied on the fact that the subtitles were burnt into the video. So we did a proof of concept and used optical character resolution (OCR) to extra the subtitles at the appropriate time. The results of the proof of concept were not very accurate because of the limitations of OCR on video but it did demonstrate capability, just in case. Luckily the 3rd party software house delivered their component quite quickly and, after some pain, the internal software teams followed suit. We ended up with the preferred solution but, given a small investment, I also had an alternative approach the preferred one had failed to materialise.

Watch the Basket

I recently discovered that Mark Twain had a tongue in cheek response to the proverb:

Put all your eggs in one basket — and watch that basket!

MARK TWAIN, The Tragedy of Pudd’nhead Wilson

This really struck a cord with me. Superficially Twain is advocating the opposite to the proverb. One proposes having more than one basket; the other says have one.
I don’t quite look at it like that. To me they are both advocating sensible risk management strategies … it is just that the strategies differ.

The Twain approach to risk is particularly appropriate in a situation where there are constrained resources. The constrained resource will often be money. Alternatives and contingencies will cost and if there are not enough funds you’ll have to do without those alternatives. For example, there might be sufficient money for one project but not enough money to invest in a separate project to develop a different approach. So you would run the single project but be diligent in ensuring its success.

The resource in question might simply be your energy rather than money. Managing several approaches (baskets) takes your time and energy. That also means you are not focussed on the main event, and with your eye off the ball, something bad might happen. Twain is saying keep your eye on the ball.

References

Grove, A. S. (1999). Only the Paranoid Survive: How to Exploit the Crisis Points That Challenge Every Company. Crown Business.