How do we go faster? A question both F1 teams and CTOs ask themselves every year. And similarly to very powerful V12 engines, small design differences in the way you set up your engineering organization can create huge deltas in your output vs the competition. This is especially true for large enterprises, where every day a new app or feature is stuck in staging and not pushed to customers, millions of dollars are burnt on expensive engineering teams that can’t move fast enough.
The DevOps benchmarking report, which measured performance across over one thousand engineering orgs on best practices and standards like DORA, shows that less than 25% of teams surveyed can be considered high-performing. Top performers reported a lead time (the time between code commit and deployment to production) of minutes, compared to most organizations which have a lead time of days, weeks, or even months. It’s easy to see how quickly top-performing companies can gobble up market share from slower peers.
In a time of inflation and rising costs across both salaries and cloud spend, ensuring your increasingly expensive engineering organization is maximizing output and efficiently delivering software marks more and more the difference between winning and losing in an industry. How do you avoid ending up paying higher and higher salaries for the same (or worse) performance? What can you do to drastically reverse this trend and cut your time to market (TTM) to gain a decisive edge?
Go faster with platform engineering
There are many different moving parts to any engineering organization, particularly at the enterprise scale. But there are really two key stakeholders you want to focus on when optimizing for better DORA metrics and faster TTM. Application developers and Ops teams.
Developers are the lifeblood of your engineering effort (and in more and more cases of your entire product or service). Unblocking them and improving developer experience (DevEx) is what separates organizations that move fast and attract talent from everyone else. The key here is to shield them from the increasing complexity of your legacy and cloud-native toolchain. You want to remove cognitive load while enabling developer self-service. This will unlock developer productivity and let you ship faster.
The flipside of enabling developer self-service is removing Ops bottlenecks and killing ticket ops. Let your Ops teams design clear golden paths for the rest of the organization. Standardize the way developers consume workflows and configurations. This will also have the great side effect of lowering your Ops to developers ratio and cutting your cloud bill.
So how do you do all of that? You guessed it, platform engineering. That’s what 93% (!) of the top performers we mentioned earlier have in common. They all built an Internal Developer Platform (IDP) that allowed them to enable developer self-service, killing waiting times and Ops bottlenecks.
An enterprise-grade IDP establishes a clear separation of concerns between developers and Ops, providing developers with clear golden paths to consume infrastructure and resources on demand, without having to wait on Ops. This massively improves team collaboration and overall morale as a result.
Adopting platform engineering to improve DevEx has a huge impact across many areas, from developer productivity and retention, all the way to TTM, revenue growth, and profitability.
Bechtle (one of the largest European IT companies) saw a 78% drop in Ops overhead and a 32% decrease in lead time and TTM after rolling out an IDP built with the Humanitec Platform Orchestrator. Convera (Western Union spin-off) accelerated their modernization journey by building an IDP, to deliver next-gen fintech products. An adoption rate of 98.8% shows developers love it, and they managed to lower their change failure rate to less than <5%, which significantly impacts delivery times.
Where to start
The key to platform engineering is to start small and prove value quickly. The graveyard of platform initiatives that died before they even got going is significant and growing. Don’t spend two years designing and planning something in a vacuum that most of your developers don’t even want. And please don’t start by slapping a portal on top of your setup thinking that will fix all problems.
Start small and quickly show value to all key stakeholders (app developers, Ops teams, execs). Follow proven frameworks to build a Minimum Viable Platform (MVP) in weeks instead of months or years. Then iterate. Listen to developers, find the right level of abstraction (without removing too much context), and prove success on important metrics with a pioneer team. Then expand from there to roll out a full-blown enterprise-grade IDP across your entire engineering organization.
Platform engineering is increasingly becoming the competitive edge of top-performing teams and enterprises. It’s what sets the Red Bulls and Ferraris apart from the rest of the pack. Get started on your MVP today and get your org moving (fast).
