Back

How Universal Banks Assess Software Developers

leave a reply | February 11 2015

BlueOptima Executive Breakfast, London – Capital Markets, 5th February 2015

To present BlueOptima's offering, we hosted an Executive Breakfast in the City of London. It was a pleasure to welcome a dozen delegates from Universal Banks - mainly CIOs and CTOs - and focus on software development best practice in this vertical. Sharing thoughts and experience led to mutually-beneficial insight for attendees.

Gherkin image

James Governor, analyst and founder of RedMonk kicked off the event with a keynote speech about the current software development "state of the nation". A focus on the popularity usage of StackOverflow and GitHub sparked the discussion around how to best recruit and reward developers. Logs of changes requested via GitHub are currently used as a basis for paying and promoting developers. Given that each element of coding work, such as a bug fix or new step in the project, requires a different amount of work from developers, this is not a fair means for assessing staff performance. This is why companies are instead relying on BlueOptima Coding Effort Analytics; a consistent, objective way to compare productivity between developers and teams of any size, methodology and technology.

The most readily understood application of BlueOptima’s offering is measuring productivity of individual software developers. Delegates questioned the usefulness of this. However, its importance can’t be understated, especially for organisations employing thousands of software developers, as represented at our event. These software developers have a wide range of abilities. BlueOptima finds that the best developers (top 20%) are typically four times more productive than team average. Picturing the distribution of this range, there need be no limit to how many times over 'rockstars' outperform average developers. Further, some developers are able to solve problems no other team members can. As the sample of software developers becomes larger, the ratio tends towards infinity – and identifying your exceptionally able developers becomes more important.

The above scenario mentions employees, but BlueOptima does not limit transparency over software developers to in-house teams. Insight into their productivity can bridge gaps in communication in terms of third parties supplying limited information about their teams’ progress. The third party may not be contractually obliged to supply all the information the client wants. They may manipulate it. Indeed, the outsourcee may only be partial to limited information due to a lack of systems for gathering KPI data in place. The granular view of software developers’ performance that BlueOptima affords has proven to reduce outsourcing costs by up to 20%, as clients are able choose which team members work on their projects.

It was suggested that Coding Effort data might spark friendly competition between developers and teams, in the way that companies have been known to publish travel expenses, who has caused and fixed outages, and even contractors’ daily rates. However, given that the data - and the very fact that your company is gathering and using it - are sensitive areas, extreme care and caution are advisable.

Even this objective Coding Effort measure has limitations: Comparable to a measure of how far one has moved through an amount of time, one can measure speed of traveling from A to B. But was this the best way to travel? Was it the best route from A to B - and is it advisable to go to B at all?! There was a consensus that Coding Effort shouldn’t be a stand-alone KPI. Acknowledging this is important in understanding that this is not a shortfalling or fallacy of Coding Effort analysis. Coding Effort is an objective metric to accompany other information that ought to be at the disposal of team leads and managers relating to whether the software in production is the best approach to fulfilling a requirement while not creating problems.

We discussed the difficulty of predicting how many man-hours a project requires. Inaccurate budgeting for software releases threatens confident execution of future projects. There may be no developer resource available for fixing bugs required for shipping a release, meaning technical debt lingers indefinitely.

Is it better to outsource software development? This can truly be a $64M question for large enterprises; and BlueOptima provides answers as to whether cheaper staff really deliver better value-for-money.

The areas of discussion above indicate the importance of BlueOptima’s reliability as a metric on which to base important decisions, from staff promotion to outsourcing. But how can we place so much faith in BlueOptima? Case studies and client pilots provide a POC for the value BlueOptima provides, in conjunction with companies' existing processes.

For the notes from this seminar [PDF, 2pp], please complete the below form. (N.B. we will not email you frequently or pass any data to any third parties).

Comments
No comments yet. Be the first.
Archives