 My name is Steve Leving, I'm a tech lead at Open Software. I'm also a certified Drupal developer. I've worked exclusively with Drupal for the last five years. I'm also a certified scrum master, certified scrum product owner, and typically I work with the sales team at Open. I want to work on projects and bids with people like the government of Bermuda. This is my fourth DrupalCon by my first time presenting, so thank you. Yes, absolutely. So like I said, I work at Open Software, we're located in Ottawa, Ontario, Canada. We were founded in 2010, we currently have about 25 employees, and within the 25 employees we have two development teams, and my development team is the one that got to work with the government of Bermuda. Okay, can everybody hear us okay? Is that me better? Okay, right. Good afternoon, my name is Martin Walsh. I'm the acting director of the department of eGovernment in Bermuda, for the government of Bermuda. I've been working there for about eight years, and I have about 30 years experience in systems. Now, just a quick question for you to get this going. Hands up, who started in coding COBOL like me? Not a lot of hands showing there, so now that could be a problem, but that problem isn't for me, so let's try another one. Hands up, who has spent, and we'll check it the inverse this way, who has not spent at least 10 years using a waterfall type methodology? Not spent, who's not spent at least 10 years doing waterfall type methodologies? One, two, three, four, five, keep your hands, keep your hands right up, right up, right up, and out of those people, if you're not involved in sales in any way, put your hand down. Okay, so we've only got one person who's involved in sales potentially. You're not involved in sales because you're in a big company, or because you're in your purely back office coders type technical people, or you're not involved in small companies, because everyone involved in small companies in sales really. Okay, all right, well, what I'm trying to do in the session is I'll try and give you a few tips of how to sell to people like me who've been around for a long time and sell them something as new as open source and potentially as worrying to people who've come up in a traditional waterfall type environment where you do everything in slow, steady stages. And a tip for you, and the tip is drop the R when selling it to IT vets like myself. Now, I'm not going to tell you what drop the R is. It isn't drop it at a Drupal, you don't say Dupal. It's drop the R somewhere else, but if you stay awake long enough in the presentation, you'll find out as I get to it later. Bermuda, just in case anyone didn't know where Bermuda was, we're stuck there in the Atlantic. We're actually the second most remote populated island in the world. There's one in the Pacific, which is further away. We're 850 miles from the US coast, Carolina. We've got a population of about 60,000 people. I work for the government and the government employs about 12,000 of those people. And now, Steve. Thanks guys. So just for the agenda for this presentation, some run through some items that we're going to talk about. So I'm going to get Martin back here in a second to talk about some intro and background to the project as we took it on. Then I'm going to talk about the response from open software. So as part of the proposal process, what we bid on the project to Martin and his team. Then we're going to discuss the selection process, some of the planning and build processes, and then close off with kind of what's next in the project and some conclusions about lessons learned. All right then. Why did we look at replacing the portal that we had already? There's a lot of things on that list which are probably common to many other people. We had an old portal. We had out-of-date content. It wasn't very user friendly. It had a poor search facility. It was unreliable. It lacked digital services of any sort really. And it was costly in the maintenance because it was Oracle. And that was one of the problems actually is it wasn't Oracle when we bought it. We bought it as a Plum Tree portal. And then BEA bought Plum Tree. And then Oracle bought BEA. And so we had this old legacy system that Oracle didn't actually know anything about, which they weren't developing. They were just waiting for us to upgrade to one of their new Oracle products and charge us a lot more money, which we didn't want to do. And we found out actually when we were taking it down that it had never actually been configured properly in the first place, which is one of the reasons it was unreliable. So many departments in fact had gone ahead and left the portal and gone their own way, built their own websites. And really this wasn't the platform we needed to launch digital services because we wanted to launch digital services to reduce costs, increase efficiency and improve service. And this wasn't the platform to do it. So when we started the project, we had three strategic customer facing goals. The first was to become a trusted source of information. We wanted the portal to be trusted by people and to know the information was up to date and what they wanted. We wanted it to be convenient. We wanted it to be available on mobile devices. The old portal certainly wasn't. We wanted to engage the public and business. We wanted to meet all the different clients. We had four strategic operational goals. We wanted it to be self-maintaining. That doesn't mean we wanted it to have artificial intelligence and work out its own content and update itself. But we in the e-government department and the IT department didn't want to be the ones having to update it. We wanted it to be sophisticated enough to let all the users update it themselves but have controls in place and governance, etc. We wanted analytics. We wanted surveys. We wanted it to be measured. And we wanted it to facilitate digital services. And that meant it had to be able to connect to back-end systems to do that with interfaces. And our current portal couldn't do that. Thanks, Martin. So I'm going to talk a little bit about Open's response and our bid process to the government of Bermuda. So looking back on the project and looking back throughout that process, I wanted to bring kind of three key points to share with you guys today. And they might not be surprising for most of the process for the proposal. First being we had to sell Drupal. So we had to sell Drupal over things like SharePoint and Sitecore. We had to sell us as open. We had to sell the team to the government. We also had to sell the fact that we weren't on the island. So we weren't from Bermuda. We are from Canada. So that's something else that we needed to make sure that we brought out in the proposal process. So in terms of selling Drupal, three key points that we had for that aspect. First being community and existing technology. So we actually helped sell Drupal by selling you guys and selling this event and selling all the people out there and all the other sessions that contribute to Drupal and contribute to the contributed modules of Drupal. And it actually went a long way with the government to make sure that they knew that there was a community out there. It wasn't just one company. It wasn't just Microsoft or the next company with its employees. It was a community all over the world. And odds are there's people in Bermuda or surrounding that also would be potentially part of that community. So it definitely went a long way. The next being government adoption. The fact that Bermuda wouldn't be alone. So making sure that throughout the process the government Bermuda can compare Drupal to things like the proprietary systems to know that Drupal is in other governments and that Bermuda if they adopted Drupal wouldn't be alone. They're not going to be the lone wolf and the only government in the world to adopt Drupal. There's many around the world. So we know that White House adopted Drupal. We also know that the Canadian government where I'm from in Ottawa adopted Drupal early on and has many departments running on Drupal. And then most recently we know that the government of Australia switched all their systems to Drupal as well. So we want to make sure that that was a selling point into the forefront. The third for selling Drupal and probably no surprise is the price licensing fees and recurring costs or the lock thereof. So the price is free and there's no licensing fees. And besides essentially professional services or hosting there's not necessarily any recurring costs to it either. Like Martin was saying before about Oracle every time they wanted to make a change they needed to renegotiate a contract and that meant a few more licensing fees and that went so on and so on and so on. So to make sure that we were at the forefront in the proposal process we wanted to make sure that it was well aware that open source truly has no costs in terms of price and licensing fees. So moving on to selling open so selling us there's three main items that we wanted to look at in the proposal as well. So that's experience expertise and the fact that we were looking for a partnership. So in terms of experience with working with other governments we wanted to make sure that the government of Bermuda knew about that experience. We've been working with the Canadian federal government and municipal governments in the area for years. So obviously bidding on this project and submitting a proposal we wanted to make sure that our expertise was at the forefront that leads to expertise. So we wanted to make sure that open was at the top of the list when it came to expertise. Open, all of our developers are certified. We actually have some operational team members that are certified. We have all the way from site builder up to grandmaster when it comes to developers. We also wanted to make sure that we promoted our standards and best practices. So when they were looking for a vendor whether it be proprietary or open source we wanted to make sure to drive home that expertise. The last item for selling open that I wanted to discuss was partnership. So we wanted to make sure that the government of Bermuda did not look at us just as another vendor or look at it as just a standard client-vendor relationship. We definitely wanted to be able to promote a partnership and a true partnership making sure that we were collaborative with the government making sure that they knew that we were there with them. So even though we had our team in Canada they had their team in Bermuda together essentially we were one project team. We also know that in terms of the partnership it leads to the collaboration aspects and things like that. And that leads to selling off-island. So having to sell the fact that we weren't in Bermuda we weren't down the street they couldn't see us all the time. So three items that we wanted to come across for that was check-ins, approach and travel. So check-ins in the sense of making sure that Bermuda knew that we were always a phone call away. We didn't have to wait for milestone meetings. We didn't have to wait for stakeholder meetings. We could pick up the phone at any time. Have a video conference at any time and make sure that we really drove that home. So it can be a random Tuesday morning or it could have been at the end of the week just to make sure we always had those check-ins to make sure that even though that we weren't on the island we can still communicate effectively. And then our approach. So our approach being collaboration was one of the key items. Even though we were in Canada and they were on island we needed to make sure that we were collaborative and we took that approach right from the start ensuring that they truly knew that and then travel. So making sure that right from the kickoff meetings, right from the initial sales process obviously making the government aware that we were willing to travel and obviously being in Canada in January having to travel to the island in the frigid cold is not a bad thing. We didn't have to twist many people's arms. I'll pass it back to Martin. Alright. So the selection process we have a quite a rigid RFP process that the government like most governments do and we came up with quite a comprehensive RFP. We actually got only 13 responses which in a way was good because we were worried we might get 100 responses because a lot of people want to come to Bermuda to make a sales pitch even if they've got no chance of getting in the job. And a quick analysis of where they came from. Six of the responses came from Bermuda themselves and all bar one involved it was just Bermuda company front-ending a foreign partner so it was really just because part of our selection process is 10% has to go towards who's got the most involvement with the Bermuda company. Three were from Canada two were from the US one from India and one from the Bahamas. So, how did open source win the bid process and this is really where I'm just going to go into a little bit of detail of really the history of IT and really the history of my career in 30 years and how that's led to where we stand now with open source. You can see in that little diagram it's very much a green amber red standard green's good red's bad amber somewhere in the middle and bright green is the best but if you look at the left hand column this is really the phases we've gone through in terms of systems over the last 30 odd years. When I started in systems everything was built. You built the system. You wanted a system you built it yourself. Then it went into a big buy phase there was lots of packages available so you got there and you buy a system. So it was the classic buy or build decision. Then in the last 10 years you got a couple of other things came on. There was a subscribe with the internet and software as a service you could subscribe to a service and you didn't have to buy anything. You just subscribed to the service you needed and then open source as well came along so open source meant that you could benefit from a number of things of other people building help building the system for you. So just looking at the build to start with and going along the top the great thing about a build obviously is you own it. That's the bright green there. When you build something you own it. You own it completely and the functional flexibility is pretty good because it's all in your control. But if you have to build everything you want the cost it's going to be a lot takes a lot of effort building everything from scratch. So the cost is high. I'm not telling you what RTW stands for. I want you to try and work that out and there's another quiz for you. There's two things you're thinking about now so no falling asleep in the audience please. So the speed to market is okay. It's in your control but typically if you build something from scratch the speed to market is not going to be great. So that's an amber and support staff and technical flexibility you need more and more support stuff the more things you build and you have to support it all yourself the more and more stuff you need technical flexibility if you start changing building on different machines databases different this that's more and more stuff so it just gets more complicated that's another red. So what happened was that was how I was introduced to systems I was involved in building a system we took it out to a bank in Egypt we supported that system and then I moved on to Bermuda I moved to Bermuda from Britain and I worked at the Bank of Bermuda and that was an NCR mainframe it all had its own systems everything was built and while I was there they went through the revolution like many other companies and countries have done and decided we cannot afford we haven't got enough staff we can't keep up with the demand from the user community for the systems they want so let's buy some instead so we replaced all the NCR mainframes we had deck facts clusters etc and we bought package after package after package and had to interface them and the great thing that gave us was the one green we got was speed to market we could deliver much quicker than trying to build everything ourselves so that was brilliant and also all the reds disappeared the cost wasn't as much because all that money building hiring all those people to build all the systems ourselves was better if we actually went the package route the support staff we didn't need as many because you had the package came with vendor support etc technical flexibility you could actually flip and flop and have different technologies involved but we also lost the greens because now we didn't own the software anymore now as a package we weren't in control of it and we were at the we were at the sort of up to the vendor to decide well what direction they're moved in we can try and influence them but perhaps we're one of many users perhaps most of the other users want to go in a different direction so we really lose control and obviously our functional flexibility goes with that we've got flexibility we can kick it out and get another package but that's another major exercise to do that so this was part of the revolution that led to subscribe and open source so where did that what did that give us well I left the Bank of Bermuda I joined cable and wireless in Bermuda and I actually left IT directly I became head of VP at Bank of Bermuda before in the IT department but now I became head of product development at cable and wireless they were trying to develop a new e-business I caught the internet bug and I was really into that this is in the in the 2000s and we were developing our own internet payment gateway the cable and wireless head office in London was and that was to be used by all the cable and wireless offices so we were then about four weeks ready to go live and the dot com bubble burst and with the bursting of the bubble the vendor that cable Mars London had employed to develop the system burst with it and went bust so we were suddenly in a mad panic oh my God we meant to go live in four weeks we haven't got a system now what do we do should we go out and buy a package oh my God no no we haven't got time for that so what we did was before the software as a service and service subscribe model became really popular we actually started doing that by subscribing New York that was white labeled we just white labeled it stuck it out there with our name as it was our system we just developed no one knew the difference it worked tremendously well we were live in weeks not in months not in years weeks so the speed to market goes to bright green for the service the support staff we didn't need any support staff it was a service they did all that technical flexibility fine you want to go the different technology we just stopped subscribing to that service subscribe to a new service so this was great but ownership had now gone all the way from green down to amber down to red because we didn't own this at all and it's really it's basically like Google mail I don't like the way Google mail works I tell you what I'll call them up and ask them to modify something it's a service they're not going to modify it they couldn't care what you want if you don't like it you go somewhere else so there's some functional flexibility you got the flexibility the cost I didn't give a green and the only reason for that is because it looks great to start with and you start subscribing to a service but for the first year the second year that's terrific but come the third, fourth, fifth, tenth year you probably find you end up paying more money subscribing to the service than if you bought or developed the system yourself so it's only an amber for the cost well then I left cable and wireless I ended up at government and then when I've been at government for a few years I became the acting director of the department of e-government and we were faced with the replacing the portal so open source I certainly heard of open source but I was not I was not a converted open source like most of the people probably at this conference are and I looked into it and obviously open and others really sold me on various things of it and the more I looked at it the more fantastic it was ownership one of the biggest problems that had gone downhill in every stage going from build to buy to subscribe suddenly came back to bright green we own this code we own this product that's terrific functional flexibility was suddenly now even better than when we built the stuff ourselves because when we built the stuff ourselves in cobalt and whatever we used modules we used calling modules we reused code we cut and pasted into things but it was our code and perhaps it was everybody else's code who worked in the same company but now through open source I've suddenly got coders all around the world who are donating modules and I've got a vendor who's got all this stuff this is like it's like cobalt on steroids you know suddenly everyone's contributing to it terrific cost low cost no license costs not nearly as bad as the packages not even the high cost of subscription but in the speed to market it's not as fast necessarily any faster than buying a package and it's probably the same support staff and the same technical feasibility so we went in there so where did that lead us? well I go back to the if you remember there was 13 people who bid for this and this way I've sliced them and diced them a bit differently we had all four of those layers we had build one person said or one company said they build it for us so we had one of those we had six buys the three in the middle there SharePoint three SharePoint two web sphere one site core they said it's a great package it answers all your needs by this office we had one which was actually hosting only which is a subscription service they said get your package from somewhere else but host it with us and we actually had a hosting aspect which was as a service from several of the others in there and we finally had four open source solutions which came in so we had all three layers in there on my diagram so sorry all four layers so what do we decide? well basically we decided to follow the Yellowrick Road except in this case it wasn't the Yellowrick Road it was the Bright Green Brick Road so we went with open for the software which was great green green green green good stuff and we went with Acquia for the hosting in a software platform as a service model green green green green that was great so now it's testing to see if you've been awake and you've had your mind working even though it's late in the evening RTW what did RTW stand for because that got a green for both any ideas? this has been the bane of my life for 30 years as a systems professional in various situations I've been forever reinventing the wheel and all those methodologies before meant that I was reinventing the wheel even when I meant from build to buy at the Bank of Emuda we kept buying all these packages and suddenly when we built everything we had one central information for one customer information file one GL one central set of codes now we bought these five, six different packages they all came with a GL they all came with a customer information file they all came with their own codes we had to interface them all we were reinventing the wheel all the time every time we changed something suddenly we don't need to reinvent the wheel because perhaps no one in the world needs to reinvent the wheel if somebody somewhere in the world can write the best system or the best modules in Drupal for this and then they can all be processed into this same you can get hold of them through a community of interest so the other if you remember back at the beginning I asked you one more question drop the R can anyone see where I drop the R that man gets the prize free trip to Bermuda just kidding that's right if you're selling this to somebody in my age group who spent the last 30 years of their life working their way through all these problems trying to not reinvent the wheel trying to get over them the last thing they want is somebody to come in and say guess what I've got a 25 year old outside who's got the answer to all your problems and all the things that bugged you for the last 30 years and it'll only cost you 100,000 shall we sign up I've heard it all before I've had so many enterprise packages tell me all that I ever needed to buy their one package and it does everything for me you don't hear that so selling it as a revolution is not the way to do it if it's somebody from my era who's used to structured waterfall technologies moving slowly if it's a government, if it's a bank just sell it as an evolution because this is an evolution as I said this is the same as COBOL was on the scale on steroids so why did Open and Acre win as mentioned already many government clients I was really impressed by the fact that Open had so many government clients in Canada Prime Minister's office the Canadian Transport Agency there was an extensive modules and community there's an extensive library out there the community of interest for Canada was very strong the community of interest from Australia that's what I'm trying to work my way into at the moment through Gartner, through Acre we got the contacts there because Acre are the same hosting company the impressive staff and the partnerships I was very impressed with Open because there were other contesters there was four people at bid with Drupal they were the right size they were young they were lean, they were hungry and they were the right size to fit the inclination for a partnership and I wanted a partnership I didn't want a client-vendor relationship I didn't want some swish salesmen to come in from IBM as they did and give me a 10-minute sales speech expect me to sign away to millions of dollars and then I'd never see them again and we just won a little dot we're not important to them I wanted a small company where we're as important to them as we're important to them as they were to us okay it was also low cost from within budget that's fine, that's very good it didn't win everything though for instance it didn't win the out-of-the-box analytics Sitecore came in they were the best presentation Sitecore blew us away fantastic presentation really impressive product but you have to make an overall choice in this and Acre themselves actually came later because Acre had already been bid by one of the other Drupal companies but we looked at Acre and we thought fantastic really like the company, really like the security aspects really liked what they got but they were only in the US we're in an offshore jurisdiction in Bermuda the government of Bermuda doesn't really want his business sitting in the US so we discounted it for that reason while the selection process was going on Acre launched their global product which made it available Open came and told us that Acre was their partner and said why don't you go with Acre it took us about 30 seconds to say good idea, we'll go with Acre and that was great because now we host in the European Cloud for Acre okay some quick figures this came in from talking with consultants we had consultants help us with the RFP etc and we were expecting a $1.9 million project over multi years to build the digital services starting with the new portal and we actually had the cost of the anticipated cost of the portal about a million dollars what did we spend what did I have as a budget when it came down to it for that year I was told everything got cut cut cut I only had $500,000 as a budget for it so that made things pretty difficult what did we get as the offers who were the final two that came down to open one out of the Drupal bids and the other bid that was really impressive was the site core that came through a local company and a US building company on cost there was absolutely no competition the site core offer over five years cost three times what the open solution was on Drupal in fact the open solution on Drupal came in at half the cost that we were currently paying in support etc to Oracle so yes it was a no brainer from a price point of view but that is not the way to sell it necessarily to the internal people in the organization who are experienced, stayed used to structured careful methodology the way to sell it is sell them first on the functionality and all the benefits it brings you and then if you show them this afterwards this is like the icing on the cake and it's like well oh yeah buy it because you've already established that this is as good as as good as or nearly as good as the product as you could possibly get elsewhere and now over to Steve thanks Martin so I'm going to talk a little bit about the planning and the build itself so yay open one but now what do we do so we went through this process like Martin said we won the bid and we have to get started so in doing so touching on a few things and we started with some project management methodologies to start with so based on the bid and the proposal that we talked about earlier collaboration and approach were some major things that we were considering and part of the reason why we were able to work on the project so right away we looked at collaboration we looked at timelines we looked at budgets and that all pointed us to agile and scrum more specifically to make sure we had that iterative approach to development making sure that we can meet some of those timelines or all of the timelines stay on budget but also include the Bermuda team and make sure that since we had short sprints we were able to get every part of the project moving forward moving along with just those methodologies as prioritization so we also looked at some prioritization techniques along with this since like Martin said we had that limited budget now we also had a tight timeline we had to look at prioritizing what actually needed to get done so through discovery sessions through a lot of conversation through some architecture of Drupal we came up with a list of functionality and essentially listed them in priority order making sure that at the end of the day we were launching on April 1st no matter what so if we got close to that deadline and we weren't done everything at least we knew the things at the bottom of that list we can go live with they weren't deal breakers all the deal breakers, all the must haves all the time and again it brought it back to the approach of collaboration to ensure that the Bermuda team was part of that process and part of deciding what the prioritization was so knowing it's not a technical presentation I still wanted to go through a few things of how we won the bid and what we decided to do afterwards so Drupal obviously was the choice as we decided or discussed but we actually went with Drupal 7 and it's contributed modules rather than Drupal 8 so when we started the project Drupal 8 had just been released and we know when it just got released there wasn't a whole lot of contributed modules in order to meet the budgets and to meet the timelines we needed to make sure that there was very little custom code or custom implementations as we moved it forward so we made the decision to stick to Drupal 8 even though Drupal 8 was released to help aid in that though when we were architecting the system we actually broke some of the portal up in pieces and used things like multi-site to make sure that we can segment aspects of the portal meaning that when the time came to upgrade to Drupal 8 we don't have to upgrade the whole entire thing at once we can actually segment the upgrade path as well and choose various items so even though we picked Drupal 7 we still had a plan for Drupal 8 going forward again not wanting to get into too much detail on some of the technical but knowing that we didn't want to produce tons of custom code or custom implementations and working with the Bermuda team and checking off as many boxes as we can for the build itself obviously we had to pick a series of modules and items to move forward with so some key security ones obviously security was a big aspect of it some key security modules that were chosen were security kit secure pages, password policy, honeypot all things that are available from the community that was so important to Bermuda to allow us to implement in the project same with some UX modules or themes making sure that we implemented navbar style guide at minimal again already contributed modules that allowed us to build out the system easily some assets so as you can imagine with government there's a lot of pages, there's a lot of links there's a lot of documents, there's a lot of PDFs making sure that we were able to still achieve that with minimal effort making sure that we were able to install modules such as media and link it to achieve some of those things to make sure that we can implement all the links and all the PDFs that they wanted without the box solutions solar being the next one so the search in the old portal didn't work very well and we knew that search was a big aspect going forward with the new portal so Apache Solar was the obvious choice it has its integrations with Drupal already and it's been proven so we went ahead with solar and then last but not least for the technical items I wanted to discuss was just the Aquia Cloud Enterprise so as Martin mentioned it was pitched before it wasn't decided on until later on but we did decide to go with Aquia in their UK data center there's a few reasons why Aquia was great and to bring it back to Bermuda right out of the box it had its team management system so we can make sure that we had the open team and the Bermuda team accessing the same system it has its workflows we can create new environments that the Bermuda team can work on at the same time as the open team and vice versa so right out of the box it kind of hit the mark for all the approach and the collaboration that we wanted to achieve as well as the technical side of things so I wanted to share this quote with you guys so open provided an excellent service and the Bermuda team is extremely proud of the end product we have a world class website that's under budget so I don't mean to share this as kind of a shameless plug for open obviously we're very proud of the work that we did with Bermuda but I wanted to share this specifically for Drupal in general so I think we can also read this in terms of Drupal created a world class website and the implementation of Drupal came in on time and under budget so just wanted to make sure that we can tie that back to Drupal and the proprietary systems that Martin was talking about earlier that would pitch that we're world class and that you just need to install it so we're going to be on time our approach in using open source and Drupal can still achieve those exact same things and back to Martin this is just the last section now it's just looking forward and what's next basically what we've got already is the portal the portal is up there and it's working it's investing online services so you can go into the portal and the customer doesn't know it but when they start clicking on pay my tax bill etc that brings them into the tax system and so forth or the immigration system for the forms but gradually we'll start launching new services but we have got a two-way communication channel because we've got feedback forms on every page in the right hand side so we can get feedback on anybody about everything on that page we've also got communication capability by two-way communication offering that we only offer that when we went live to departments that already had email process in place because we didn't want to just make every department have email communication they never had before when they hadn't got the process in place to deal with it because it's nothing worse than launching a service an email that no one looks at because then you'll just you'll just get a bad reputation what's in progress we've got standard cookie-cutter forms in place e-forms using Drupal which are gradually going to replace all Acrobat forms, Woofoo forms, all the other types of forms we've got on there at the moment we've got standard e-payment gateway we're going to go up with and we've got links to online bank bill payment services in Bermuda and we're also about to embark on a pilot with MuleSoft for the Enterprise service bus using RESTful APIs that will let us link to all these legacy systems we've got AS400 systems we've got some Unix systems we've got Windows systems we don't want to replace all those but we want to be able to set the data out and use it in our digital services and then we want to be able to pass information back from the forms and feed those systems as well so that's part of our intention and we've got a standard reporting mechanism we've got good analytics we've got good surveys we've got the feedback forms we want to keep on top of all that so the conclusions the very final stage how did we do against schedule as you can see just from the top that's a very aggressive schedule that we were set the target date for and this was we launched the RFP on schedule in July and we had a very aggressive schedule the contract signed by October 2015 the hosting set up by December the build complete in February so they had five months from when the contract was signed to get the build in then a month of user acceptance testing in March and we had to go live on April 1st I'd already tried going back and saying well April 1st but you should have but later April 1st so how did we do against those schedules well we started great with the RFP or went on schedule we actually even made the choice on schedule but then when it came to getting that schedule getting that approval process through the stages within government of you know the the different sort of levels of management level and then ministry level and then cabinet level and then everything else it took to December 21st before we got the contract finalised and signed which happened to be about two and a half months later than scheduled so I trotted off back to my PS and said well this time it's actually your fault it's all your people up at the top it's your fault so obviously we could now move the date from April 1st because that's not possible anymore because look we're two months behind this impressive schedule ok yes April 1st so we didn't get any joy there at all but in came Acrea in their white cape so we didn't sign the contract till four days before Christmas fine you want your new setup live by December 31st open and Acrea together got that new setup for us in December 31st so that was in ten days including Christmas, New Year, Boxing Day all that was phenomenal to make that so the next challenge was opens we said they could have five months to develop the system I then went back to them and said well I know I said five but I really meant two and they delivered February 29th good job it was a leap year that software arrived that software arrived it then took us a couple of weeks individual testing in this time to accept it but that left us two weeks not the four weeks or in fact it ended up as seven days I think of or seven or eight business days to do the user acceptance test which is not the depth we really wanted to do it in but April 1st it went live so it was on schedule and it was on budget because I know we only said we had a 500,000 budget but I managed to supplement that budget by canceling the Oracle support contract early which cost us 230,000 a year and finding a much cheaper way of supporting what we already had the interesting case it fell over before we managed to replace it and that was one of the original plum tree staff that implemented it for us so that was charging a little bit less than 230,000 for doing nothing which is what basically Oracle did so we got there that together added up to about 700,000 we got there in 660,000 and we only had an 86,000 ongoing annual cost but of course a lot of cost to come with the digital service development and sorry I meant to point out on that last one the split if you like was a quarter of the budget went on the RFP, the requirements the RFP design which we involved some other consultants on they were there with the company we're very happy with them and they introduced us to another partner FordView who were also exceptionally good and they really helped us not only in that initial stage because they're another Canadian company but in the portal content they really helped us in rewriting that content because that was a huge task to rewrite that content so what happened in terms of satisfaction though did we meet our objectives we had some very good figures in terms of the percentage of the population that used the portal going back to right when I first joined the government in 2008 and what the satisfaction was back in 2008 when the old portal had only been live probably a year or two only 30% which was very disappointing of the population used it in our random surveys but 73% of those people who used it were satisfied with it which was great so we went we had surveys like every 3 months, 6 months or whatever but skipping forward to make a long story short to July 2011 we had doubled nearly the usage to 53% of the population but at the same time we'd halved the satisfaction because that had gone down to 34% and then you skip we had more surveys but skip another 5 years and the very last survey with the old portal in March 2016 showed that nothing had really changed in the previous 5 years and that we're now 54% not 53% which is not significant and 37% satisfaction not 34% which is not significant either it's just to be plus or minus 5% but when we launched the new portal it had been in for 2 months and I took a survey again and the results were somewhat disappointing but basically the old adage of you have to pay the piper somewhere on the line came true because we can't cut all those corners and rush and keep it in to go live in April without rushing some things and some of the things we rushed was the way we launched it and the way we involved we had involved all the users a lot in focus groups, in designing the new portal and then we went such a sprint to get it launched we hardly involved them at all we had to invite them back to let us tweak the new content and we had to make decisions to change things a little bit to get live so one day they didn't like the old portal but they were used to it and it was always the same next day it was different it was a different design they didn't like it so we had to overcome that but luckily we did overcome it because in only 3 months not as much as I wanted to and still this is a met in progress for the satisfaction but we've already got the usage up to a very impressive 66% so that's 2 thirds of all the adult population used the portal not in the last 3 months like the original figures were but in the last 2 months so we were very happy with that so we have now the platform for digital services we want to launch and the satisfaction hasn't gone up hugely yet but it's gone up to higher than it was before and that 41 we're going to push up till that's up to the 60, 70 even the 80 mark would be our target alright so what were the lessons learned resources budget and time open source cloud agile, motivated project team get those 4 things together the next major project you run and you're well set to come in on time come in on budget and come in with a limited resources as I'm sure you probably have everyone has limited resources treat your vendor like a partner and hopefully they'll treat you like a partner don't just look at the cost don't go for the lowest because open weren't the lowest the lowest bid we had was actually the life ray bid company but they didn't even make the shortlist because there's other things you need to look at so look at the vendor's record what have they done with other people but at the same time look at the people they've got now because they may have a fantastic record but all the people who built that record may have now left you've got to look at the people you're working with make sure you get the 18 we got the 18 with the implementation and we're really happy with open and what they did for us they're a massive company as well I didn't want to go with a huge company I didn't want to go with an IBM and this and the other so we're just another dot somewhere a million we're a blip in their profits I wanted to go with a small company where we meant as much to them as they meant as much to us I mentioned that earlier you can leverage their partners because we did leverage their partners it wasn't just open that helped us on this it was Acquia so who's that the one where we failed really to some degree was keep the momentum going one of the difficulties we had was coming in under budget effectively certainly under what we expected and on time and all the rest of it is well that's done and dusted let's move our focus on to something else now but it's not really done and dusted it's the portals live but we still got years of digital services and I've got an ongoing struggle like most people in government get in the budget, get in the focus get in the priority, get in the staff to move as fast as I'd like to on the digital services but you keep it going as far as you can and the other tip there the bottom is build for the future now we found that out from the survey results pretty convincingly if you remember back to the slide said what was my objectives when we went in we had a lot of problems to fix but you don't get a lot of thanks for fixing those problems so don't underestimate the work it is to raise public satisfaction and that leads into the next size that one size doesn't fit all even though we think we've got a really modern design and we're great the age group that we've put off somewhat was our biggest user group before which is the 45 to 55 year olds they were our the biggest users of the old portal and they are the least satisfied with the new ones but they're people because they're on mobile devices etc etc so our usage of the 18 to 30 year olds has rocketed up which is great and the final lesson there was how you solve a solution so it's really horses for courses remember at some stage in the authorization process there's going to be people like me people who have grown up with use to waterfall methodology particularly in governments financial companies and all that and if you're selling if you're the salesman selling to those people you've got to validate their life before and show them it's evolution it's not a revolution no one wants to hear it's a revolution no one wants to hear that I've just waited 20 years of my life doing it this way and you can solve all the problems with one little system because I've heard it all before and it doesn't serve all the problems but at the end of it you can show it's great value for money then that's a win-win all around stress it's low risk and that will clinch the deal it's established it's a low risk it's a logical thing to do but and it's cheaper and it's half the cost of the others that'll win it finally I'm just going to leave you with the portal as I haven't shown you any of it before this is what it looked like on a handheld device this is what it looked on a tablet straight away the old portal didn't look like anything on either of those and this is what it looks like on the desktop and there's the web address www.gov.bm please go on, please have a look and please give us feedback if you see that little on the you can hardly see it but on the left hand side there's a little feedback tab you can give us feedback on any page you can tell us how you think it should be improved you could improve it what we could do to make it better, what you don't understand we love it so just give us feedback I'm going to pass it back to Steve well thanks guys so that's it for us please make sure that you check out the website where the slides will be posted and the recording will be posted also make sure you take a survey tell us how we did we appreciate it sprints so we're asked to make sure that we make everybody aware and remind them of the sprints on Friday I don't know if there's any developers in the room but none the less very important ladies and gentlemen five more posts in five minutes I guess so something that we actually just released that opened a couple of days ago is a new module called content synchronization so we just wanted to share it with you guys as well essentially it helps with synchronizing content from one environment to another we face this problem all the time in every Drupal implementation so we're kind of proud to put this on Drupal.org you can definitely go check it out it just got released so it's in beta but we're definitely just looking for some feedback and things like that on it apparently we're out of time but the details are here for myself on Drupal.org if you want and Martin's email is there as well we're happy to discuss anything in the slides of the project we're here for a couple more days as well so if you see us in the halls please stop us and we're happy to chat thanks guys