Print Page   |   Sign In   |   Join

North Region Member Meeting

Workshop: Change and Release Management

21/11/2016 » 22/11/2016
ITSM16 Conference & Exhibition

Workshop: SIAM

Masterclass: Major Incident Management

  Discussion forum  
Share your views on ITSM on our Member Discussion Forum
  Social Media  
Follow us on...


Its's all down to teamwork at OVO Energy

It's all down to teamwork at OVO Energy


OVO Energy won ITSMF UK’s 2014 Award for Team of the Year, presented to the company that can best demonstrate outstanding team working and show customer commitment above and beyond the norm. So what made the OVO team stand out this year? Catherine Cheetham and Jon Dodkins provide the background to the story.


Managing rapid growth


OVO Energy is a growing business. The service management team at OVO were serving 180 full-time employees (FTEs) in January 2013, and this number had risen to 530 FTEs by July 2014. At the same time, OVO customer numbers grew from 130,000 to 340,000. Successfully supporting new (and rapidly changing existing) business functions, increased complexity, office moves and expansions with an inevitable plethora of new and legacy third party relationships has been a huge challenge for our team.

In spite of these obstacles, the team have been able to continue delivering changes and maintenance tasks at a ‘start-up pace’ of change, some weeks seeing up to 50 Requests for Change (RFCs) released into the production environment. All this whilst they were building and implementing entirely new processes, achieving a high level of customer satisfaction and high availability across core services.

In January 2013 there were 15 staff in the Service Management (IT & Systems) team. We had already begun implementing incident, problem & change management, documenting our service portfolio, service catalogue and using capacity & configuration management to prepare for what was to come. Needless to say resource was an issue; we lacked the engineers we needed along with key management roles such as a change manager. Having recently opened a new office in Bristol, new starters were flooding in as well as a massive relocation project getting staff moved from the previous HQ in Kemble, Gloucestershire.


We had to find new ways of working smarter with what we had in the face of a burgeoning workload. We couldn’t do everything we wanted to, so prioritisation was (and continues to be) massively important. We were using systems that were quickly becoming unfit for purpose and a lot of free open source tools which, despite what we gained in cost savings, did not meet our requirements for further growth.  

Everything OVO do is underpinned by IT & Systems. It was critical that Service Management maintained a good grasp on BAU activities while supporting the company’s appetite for new projects and business ventures. Between January 2013 and July 2014 OVO have formed completely separate business units for Pay As You Go, SMART metering and OVO Communities; each department with their own technology requirements, management structure and goals. 

Our team had to evolve to balance reactive tasks involving creaking systems, major incidents and last minute business demands with essential proactive planning and problem management activities to ensure that we not only maintained but continued to improve the quality and scope of our service offerings. 


Meeting key objectives


Our Service Management team decided to take an innovative approach, combining the principles of ITIL and Agile frameworks to ensure that we were able to move quickly and flexibly without compromising stability and quality of service. Within this broad approach we set out to meet three key objectives:


Reduce OVO’s cost to serve and propensity to contact through self-service.

‘MyOVO’ is our customer portal; our ‘shop window’ that demonstrates our company value proposition that managing your energy supply should be simple and zero effort. We recognised that encouraging customers to self-serve would improve gross margins and, if we got it right, reduce our ‘cost to serve’ and ‘propensity to contact’.  

MyOVO 1.0 was launched in September 2012 by a third-party web design agency, but initially we weren’t seeing the reduced call and email volumes that we expected. In 2013 a company-wide objective was set to deliver 80% of customer journeys online with an ‘OVO experience’, which is so important to our culture.

Coding started in earnest in April 2013 and MyOVO 2.0 was released in October 2013, six months after the source code was in-housed. This was a highly visible project, expected to deliver immense company benefits. Also, as a company-wide objective, everyone’s bonus depended on successful delivery. We achieved our objective by:

  • Bringing source code in-house.  There was no time for a complete rewrite so obtaining the source code from our third party was vital.  Through effective supplier management using a contract that had been very poorly written, we negotiated the full rights and handover of code.  
  • Addressing weak configuration management through parallel running. MyOVO 2.0 would be run alongside MyOvo 1.0 and customers could make a choice over what they would use. Two different configuration sets, infrastructure platforms and service designs required tight configuration management; Linux and Windows are very different beasts but using Chef we could manage them easily.
  • Addressing poor capacity planning and scalability through investment. We now have a service-based structure using specific services scalable for future product development; this follows the same principles but is based on a reliable platform.
  • Hiring a business facing, technology-savvy product owner with commercial accountability for developing a revenue generating digital product.
  • Prioritising features against problems, focusing on those likely to deliver the biggest benefit. This ensured ‘coal-face’ decision making as if the customer was in the room. We worked on the ‘big ticket’ items, both new and old, maintaining a 50-50 split between new features and support/maintenance. This defined focus meant that stability and innovation were achieved.  
  • Introducing Agile working practices which ensured that we released frequently and sought immediate feedback. The product owner and service managers spent days in the contact centre listening to our customers; this feedback loop from agents to developers proved invaluable and ensured we went to market with what the customer wanted.
  • Adopting the zero effort principle, which focused our attention on ‘right first time’.  Our release processes had to be ‘fast to market’ but with stability. No emergency releases and more normal changes. Quality was assured through developers doing their own unit testing, with testers picking up the Waterfall test phase for full regression testing and more user acceptance testing. 
  • Using a support model which reflected a DevOps philosophy, with a multi-skilled team capable of developing, monitoring, and releasing.  Service Desk continued with traditional first and second line support, with developers acting as third support.  
  • Diagnostic-led problem management using a tool called logstash to record transaction success rates within My Ovo 2.0.  One emergency fault fix release instantly fixed 872 issues out of 883. 

This was just the start… 

Before the launch of MyOVO 2.0 we estimated fewer than 20,000 logins a month. In June 2014, 473,000 logins were recorded. Between 24th and 30th June we successfully captured a 99.9% transaction success rate. In the last week of June we had 390 errors out of 431,000 unique customer actions within MyOVO 2.0. 

Propensity to contact (essentially, the number of times we are likely to receive contacts from a customer during a defined period) is measured between two points: 45 days prior to a customer’s supply start date all the way to 60 days after their contract end point; this reflects the entire period we could (in all likelihood) be contacted. In Feb 2014 this figure was 4.4, and reduced to 2.4 when measured in June 2014, supported by the full push to the new MyOVO. 

Translating this into cost to service – at 59p per minute over an average ten minute call, this has reduced our per customer figure from £25.96 to £14.16. Taking this saving as £11 across a 350,000 strong customer base equates to a massive saving of £3,850,000 annually.


Achieve 99.9% availability across customer-facing services

Availability management is not a high priority for a start-up; we weren’t really interested in anything outside of 09:30 and 17:00.  But with a Q4 2013 strategy for high customer acquisition, we had to focus on measuring customer facing services and guaranteeing ‘an OVO customer’s ability to use our services’.   

We had no real-time reporting. Our only data was from an in-house developed transaction processing monitor using the personal OVO Energy customer account of one of the contact centre agents. So using service outage analysis we identified that the greatest strain was being felt on our most important system, Gentrack - our billing platform. As this was the first system to be installed for OVO, we had developed it quickly and cheaply, resulting in a service design that did not do demand management and did not support extreme peaks of activity and throughput. After continuous customisation, Gentrack has now been heavily modified and architecturally evolved to master all customer data, and therefore supports all customer-facing systems through APIs.  Our digression from base configuration and hyper-growth has resulted in a complex service to manage but the stability of the Gentrack platform was key to achieving the objective. 

Using a combination of tools - New Relic for front-end availability with logstash and kibana for user journey performance - we could quickly correlate exceptions and unexpected errors. We were not going to focus on overall availability but error reporting, where each error was managed as a problem and worked into iterative development sprints for investigation and resolution. Through extensive problem management, strict CSIP processes that clearly articulated benefit, risk and priority with determined supplier management, we were able to recover from any downtime experienced within our SLA and, more importantly, we hit our objective of 99.9% availability in December 2013.

This achievement should not be understated. In October 2013 our brand awareness was just 8% (based on independent surveys). At the end of October, after OVO Energy CEO Stephen Fitzpatrick appeared at an MP Select Committee to hold the Big 6 energy suppliers accountable for their seemingly inexplicable price rises, our brand awareness shot up to 25% and created unprecedented contact for OVO, in both the Contact Centre and visits to the website. Our first advertising campaign ‘Feel Loved Again’ launched in February 2014 and again saw huge spikes in our customer activity.

To continue tracking, we now have weekly reports on all sites measuring downtime, uptime, mean time between failure, duration and frequency, providing us with historical as well as week-to-week performance. All tools cost money, whether off the shelf or developed in-house, but the investment is certainly worthwhile.

improve scalability of technology delivery and operations capability for OVO Energy.


A company-wide initiative, ‘Ready to grow’, prompted technology to focus on demand management and the proactive assessment of our systems and environment for the next big growth spurt. 

The service improvement plan included people, process and systems, with the ‘big ticket’ items delivering the biggest benefit being prioritised.

Some of the key deliverables were:

  • Change of Tenancy (COT) automation.  This was a very labour-intensive business process. Being a seasonal activity, we were unable to employ full-time staff so ended up with a large temporary staff pool and no knowledge retention. Service Deck tickets increased and SLAs were impacted. Through proactive analysis of the top five logged tickets we recognised that a new COT wizard would address a number of issues and we developed a wizard (basically a series of questions and scripts) providing knowledge and therefore eliminating error. The result saves approximately ten minutes per COT processed. Considering that we’re now processing 550 COTs a week, this is a weekly saving of over 90 hours.
  • Broker file sizing.  This concerned a key data flow from broker sites, where the quicker they were loaded the more time was allowed for other on-boarding processes, all of which are governed by OFGEM. Through database investigation, careful capacity planning and multi-streaming of processes, we reached an optimum file size of 750KB (the equivalent of 846 customers) from 305KB (420 customers) per file load, which effectively halves sales upload time.
  • New payer file efficiency – a priority process which collects the cash; the result of the process running to time can mean thousands for OVO’s cash flow.  As we have grown, we have seen a gradual degradation in process efficiency and run-time.  Being an intensive process, we were unable to run it within business hours, restricting our capacity. But with an additional database index on a scarce table, process times reduced from 9 hours to 4 hours.  
  • ‘GenExchange’ upgrade. With more customers, we had more communication and our capacity to send increased batches was deteriorating over time. By adding additional servers and streaming processes across four GenExchange servers, bill runs could be sent within 12-16 hours where previously it was taking 4-5 days to clear down.
  • Contact centre platform replacement.  A very visible and high-profile project, both internally and externally, directly impacting customer experience.  Since the very first system was replaced back in June 2012, we’d been plagued with various incidents (some of them major), including poor voice quality, configuration issues, hardware failures and an unintuitive design that led to various agent behaviour/training problems. After months of planning, requirements workshops, build and development we released a new CRM-based contact centre platform on 22nd May 2014 with great success.

Teamwork at OVO Energy

In true ITIL fashion, the benefits of OVO Energy’s approach to service management success can be broken down into the three Ps: people, process and product/toolset.


In all departments at OVO, finding the right culture fit when recruiting is by far the most important selection factor. IT & Systems are no exception. We look for personable, interesting, hard-working, passionate, humble and ambitious people to join our family.

Keeping our small start-up culture as we grow at OVO is paramount to maintaining our agility, high rate of change and high employee satisfaction. 

All new staff joining the company attend an ITS induction session that is perhaps a little different from the IT training they are used to. It’s a session where roles and responsibilities across the entire team are introduced and explained and, along with some security policy points, active discussion is encouraged around previous experiences with IT and Service Management. We’ve found this initial rapport-building vital to the team’s success in setting (not just managing) expectations. In the induction session, our ITSM toolset is also introduced and we explain who to speak to when something needs to be escalated.

We are serious about being approachable; despite massive growth and the demands that places on the team, everyone in OVO is encouraged to walk up to any member of our team (even the CTO) and ask if they are stuck or urgently need something. Our Service Management challenge is not about avoiding people, but ensuring our online self-service portal is a premium service such that only urgent queries and ideas need to be engaged in person.


Having such a small team forced us to distil all methodologies, frameworks and collective experience into a flexible, bespoke approach that suited OVO. We had to focus on what actually works, as there was simply no time to waste with excessive process, neither could we afford to risk exposure to a company-ending crisis should the requisite structure be lacking.

High workloads and ambitious timeframes will always be placed on the team in this kind of environment. OVO want to launch new ventures quickly with minimum fuss and cost. Rather than develop a resistance to this, we adapted, leveraging a clear opportunity to build real relationships and working rapport with all OVO staff. Back in January 2013, the company was certainly small enough to achieve this easily and we have placed high value on continuing to dedicate effort to this. You have to adapt to your customer base.

Achieving process buy-in is especially challenging when moving a small start-up to medium sizing. The close-knit relationships that we have built within the operation have allowed us to sell the gradual introduction of necessary process and workflow without resistance or cynicism. 

In this time period we were able to successfully implement a more mature change management workflow/process (also extending CAB to eCAB), refine our incident and major incident management such that any individual can pick up the procedure, and concretely embed problem and knowledge management into the collective OVO consciousness. 

Indeed, our ITIL problem management model has been adopted by other business areas to structure their own approach to tackling their ‘growth pains’, championed by Service Management.

The toolset

In March 2013, we rolled out BMC Footprints as our primary Service Management toolset.

Prior to this we had been using a free, open source product called Spiceworks. Whilst it addressed our needs at the time, it lacked the structure and ease of workflow customisation we needed to take our Service Management approach to the next level.

We chose the Footprints product because of the ease and depth of self-customisation available. We built a comprehensive customer self-service portal – complete with a basic service catalogue and quick ticket templates – and launched OVO Motion to much fanfare. We knew the key to making this a success would not be getting everything perfect on launch day, but rather establishing a feedback loop with our customers and an efficient mechanism to quickly turn around on-going iterative improvements to the system.

Since launch we have refined all corners of the service and continue to glean feedback from regular user groups, steadily making changes to improve customer experience without compromising the management information we need to incident match, prioritise and maintain good visibility of our environment. 

Around the same time we also installed BMC’s Asset Core product to get a better view of our desktop and server estate, refining our approach to inventory, software deployment and desktop management.

Bringing it all together

In September 2013 we launched a new intranet CMS platform on Sharepoint 2013 – called the Kitchen Table.

Back when OVO were very small, the Managing Director Stephen Fitzpatrick would have a daily meeting around his kitchen table for everyone to share what they were doing that day. This ethos has been embodied in our intranet service through its name and approach. 

Besides embracing the Facebook-meets-Twitter style interface across the business, we in Service Management specifically intended to exploit the knowledge management advantages Sharepoint brings. We created a ‘Wiki-Geeks’ user guide section that encourages users to Google their issue within the document library, picking up relevant documents via metadata keyword tagging.

We began publishing our weekly KPI Dashboard for all to see, and even creating ‘how to’ videos so that OVO staff could train themselves. Utilising the Sharepoint list/calendar views, the business weekly CAB and IT&S Technical CAB meetings became slicker and, as access was open to all in one central place, valuable time in the meetings themselves gradually became more efficiently spent in useful discussion over 

the week’s activities.

Where the RFC throughput is high and change is rapid, ‘closing the knowledge gap’ has become our most valuable support asset and enabled us to maintain a self-service service desk function without compromising the quality of interactions we share with our customers.

In conclusion

IT and Systems are very much seen as the ‘team of choice’ to work in at OVO. Any secondment opportunities advertised are met with a flood of internal applications from all over the business. The appeal no doubt stems from the obvious camaraderie of the team, the leaps and bounds we’ve come in tangibly improving systems and services, and the opportunity the team presents for both progressing one’s own career and making OVO stronger, all great reasons to climb on board. With further growth, a new office, the expansion of OVO Group, Pay As You Go, Smart Metering, OVO Communities – all wrapped up with an increasingly innovative approach to both ITIL and agile in the team… it’s an exciting time.


Jon Dodkins and Catherine Cheetham of OVO Energy collect the award for Team of the Year at ITSM15.
Sign In

Forgot your password?

Not a Member Yet?

  Service Talk Newsletter  
Download the latest Service Talk and browse recent back-issues