The Meta/FB downturn as a case of the McNamara Disease
TL;DR: A culture that obsesses over improving metrics loses the ability to disrupt.
Proper disclosure
I was employed by FB in 2018-19 as a Senior Software Engineer. I own FB stock. The opinions and observations stated here are solely my own, and are a reflection on “ambient culture”, as opposed to a hard analysis. I have never been, and I am not now, in meaningful contact with any kind of Meta/FB leadership.
Who was McNamara and what did he do that was bad?
Robert McNamara was the US Secretary of Defense during the onset of overt US military involvement in the Vietnam War. An operation-researcher-turned-senior Ford executive, McNamara championed a hard quantitative approach towards problems of policy.
Modern militaries are highly bureaucratic, and report on a rich family of metrics. Two central metrics around warfare are the number of enemy casualties and ratio of the enemy’s losses to friendly losses. The “old school” expert advice that McNamara received was building support for the US with the local population first and foremost through peaceful means; however such policies are not associated with a simple metric that could be maximized. McNamara’s instruction to his people was to focus on the easily measurable - i.e., casualty counts, and as we all know – the outcome was tragic.
What transpired is that a powerful military, such as the US, can be extremely effective in killing people - indeed, even combatants - with minor losses, too - but without actually “winning” in a useful sense of the word. A strategy that was used to obtain impressive results was targeting non-critical military facilities away from the combat zone. However, it did not improve the situation on the front-line, and indeed if anything, consolidated an operational stalemate that made American victory impossible. The real end result was an endless conflict that caused unsurmountable suffering to all sides.
A few words of background about Meta/FB
At the core of Meta/FB’s business value is a single, extremely elaborate social network product. Social networks by definition allow to gather a rich family of metrics, such as daily active users, advertisement exposures, clicks and so on. Meta/FB has a variety of side projects that are mostly strongly decoupled from the main platform – many of them are kept for externalities rather than for direct business value; I will disregard them because even the largest of them are an order of magnitude smaller than the business core.
For many years now Meta/FB’s history, the performance of individual contributors has been measured via their contribution to the metric counts associated with the product/area/stack to which they were contributing. The organization came up with highly sophisticated tools for incremental experimentation on new features and analyzing the impact of such experimentation on top-line metrics; employees were expected to perform several iterations of experimentation for every 6-month period. At the end of such a period, employees were requested to justify their employment through the metrics their work improved over these months.
Personal impressions
Being an employee at Meta/FB was a “drinking from the firehose” experience. So many voices are heard that some critics compared it to an ever-going popularity contest. However, one kind of voice that I personally found eerily absent was auteur-ship – the feeling of “here’s a major change I originated and can own, let’s go for it”; the absence was especially notable from the middle management, which is the organizational layer that determines an organization’s ability to innovate.
Indeed, any preoccupation that did not coincide with the possibility of improving metrics over at most 6 months was considered unhealthy, and people were strongly advised away from it. (I learnt that recently the reviews were switched to being yearly and not 6-months; I doubt that this is going to make a difference)
I have most definitely heard of several major product initiatives coming down from senior management; however I never saw in close proximity a ground-level person who had the enough authority and agency to be a major part of the product improvement process.
One important clarification to make is: a lot of the people I met were great and could lead a major change in product, but very few did; this does not surprise me, because there was no way of doing that within the framework of the organization; one wouldn’t want to attempt that and risk having a Not Meeting Expectations in the review, would they?
The Product Perspective
A lot of people over Hacker News (link, link) make the point that FB/Meta has found itself out-innovated in the video creation niche, which appears to be driving usage these days.
Personally as a user of the platform – a content creator who used FB/Meta to create multiple types of content over the years – I must say that it became increasingly unbearable as a product. There were so many missing features: Every possible product dilemma was twisted in the direction of creating immediate, fleeing engagement, which went against me being able to create a faithful representation of my life and my ideas. A further concern were the community dynamics – I don’t think I’ve ever seen a profound and respectful conversation within the Facebook platform, and it does not appear to me to be a major product or operational goal of the organization.
What ended up being the straw that broke my back were the bugs - text just vanishing when I was editing it (of course, that happened as the internal review season approached! I hope that the poor soul responsible at least got a decent review). I voted with my feet, and established a blog outside of Facebook; one of my medium-term plans is to add a self-hosted comments section.
The Systemic Picture
One key principle that I see all of the time is that an organization becomes what it measures. After many generations of employees and managers who were selected not for having a vision, but rather for being able to drive specific key metrics, I would be extremely surprised if the organization is still able to withhold the huge stresses and changes inherent in substantial product innovation. (I can claim to some expertise as I have been doing substantial innovation over much of my career)
As a comparison, Apple has had an fundamental pillar of product design from the very start, involving top-level visionaries such as Steve Jobs and career professionals such as Jony Ive; personally I have difficulty imagining the current kind of FB/Meta leaders tuning into a similar signal. Accordingly, a change is unlikely to occur until the leadership is replaced.
My diagnosis of FB/Meta as being affected with the childhood McNamara disease also leads me to doubt the organization’s ability to deliver radical innovation in the AR/VR domain, which is the banner currently held by the leadership. Innovative products require being filled top-to-bottom, bottom-to-top with enabled and empowered leaders of change; if anything the system currently alienates such types.
It took Microsoft a lost decade under Ballmer until the right person - Satya Nadella - came along and washed out the culture that was holding the organization back; as painful as it might be, that’s what I consider the most viable path going forward; as a shareholder, I can only hope that such growth would take place before it’s too late.