Trimantium GrowthOps (formerly 3wks)

BACKGROUND

Please provide a brief description of what your company does for some background information.

We are one of Australia’s largest media companies, spanning newspapers, magazines, online, and subscription TV. We publish over 140 online and printed newspapers in major Australian cities including Sydney, Melbourne, Brisbane, Adelaide, and Perth, as well as in suburban areas.

And what is your position within the company?

Head of commercial platforms and products.

OPPORTUNITY / CHALLENGE

What were your company’s business goals or reasons for undertaking the project?

Like many publishing organizations, we rely on classified ads from consumers, local businesses, and large enterprises as a source of income. Traditionally, booking and billing classifieds was manual and time-consuming. It was a mix of either field sales teams call centre agents, or basic online tools. Ads that are not booked online involved manually intensive processes. The potential rose for inadvertent errors such as typos or omissions. Seeking to be as customer-focused and efficient as possible, we wanted customers to be able to serve themselves by placing ads online. For customers, the classifieds booking system needed to be easy to use. It needs to give our customers more control over their ad content and allow them to publish ads online immediately. Our goals included fast time to market, lower costs, and less overhead in the form of call centre time and manual data entry.

SOLUTION

What was your process for selecting 3wks to work with?

We were looking for a partner with a start-up mentality.  Traditional technology houses couldn't cope with the degree of unknowns we started with.  Agencies couldn't help us with the heavy-lifting technologies.  Neither could move at the speed we required with all these unknowns. 3wks immediately stood out as specialists in a new field:  the art of committing to delivering to real users at lightning speed before you know what it is!

Why did you decide to build your application on Google App Engine instead of similar platforms?

We chose Platform as a Service (PaaS) over the traditional Infrastructure as a Service (IaaS) because we estimated a 40% lower project cost.  This was based on previous experience with both platforms.  With IaaS, we need system administrators, database administrators, network specialists, application server specialists, and network security people.  With PaaS, we avoid these costs (and more importantly the time delay and dependencies) because world-class resources at companies like Google have already solved all of those problems.  We effectively get them for free.

This also reduces our operating costs dramatically because PaaS is elastic, which means that we don't have to have 40 servers just because we need them on one day of the week.  We pay for what we use, which has turned out to be a massive cost reduction for us. We chose Google Cloud Platform over Amazon's Web Services because Google's PaaS offering was more mature and pre-integrated for building web and mobile apps.  They have a host of convenient, built-in services.  For example, Geo search, big data services, and the Google Prediction API.  We didn't waste time choosing and integrating these services from third parties. With the Google Cloud Platform, we started building real features for our customers on day one and had them in users' hands within three weeks of the project starting.

Can you tell me about the scope of the project?

We built five separate web applications in the classifieds genre.  Each one had a different customer base, branding, and architecture.  Some of these were deliberate throw-away market tests (micro-investments) in local geographies before we built for the national market, like Traderoo. Our project team was fully self-sufficient with a blend of marketing, technical, and business support skills.

What parts of the project was 3wks involved in?

3wks provided a fully managed service in the early stages of the project. From branding, design, and technical build through to technical support, quality management, and customer support in the early stages.  We were able to focus on the business strategy and executing our marketing plan.  This allowed us time to establish a team of world-class resources over a more controlled period.  We now fully support and extend the platform with internal resources, with peak cover from 3wks.

What areas were you able to develop in-house?

Due to the time pressures on these projects, we held to a specific strategy of avoiding changes to internal systems wherever possible.  The only changes we made internally were for system integration, which was surprisingly pain-free. We were able to leverage investment in internal technologies made in the previous year.

What technologies were used?

We choose to use Google App Engine (Java)

• Built-in task queues
• Built-in distributed cache
• Built-in page speed cache
• Built-in A/B testing facility
• Built-in "backends" and CRON jobs
• Built-in Geo IP headers
• Built-in distributed logging log query feature
• Built-in warm-up requests

• Data Store
• Geo Search on Data Store
• Search API for full text search
• Big Query
• Blob Store
• Cloud Storage
• Maps API
• Prediction API
• Google Apps
• Domain hosting
• Mail hosting

Can you give us a sense of the size of the initiative in dollar terms or person months?

Sorry can’t discuss that.

How recent was the project finished?

This is effectively a startup business within our existing enterprise.  It continues to evolve at a rapid rate and we continue to make significant investments.

RESULTS & FEEDBACK

What are the results of the project, and if you have any stats or metrics I’d be happy to hear them.

We currently see project/operational cost savings above previous levels. Over the first 12 months, we made 157 releases into production. Our largest release was 3 weeks from idea to users. This was typically the first release of a new product.  The fastest idea-to-user turnaround was five hours. Our big data platform (BigQuery) has a single data set of over 300 million rows, which took us less than ten minutes to load from around 200GB of data in the blob store.

Since building the application, how has the Google App Engine platform/service/platform performed and do you still feel you made the right decision?

We're very pleased with the Google Cloud Platform.  We've been surprised by the degree of local Google support here in Sydney.  It's fast becoming our de facto standard for a host of other solutions in the mobile/web application space.

How did 3wks perform overall?

3wks is exceptional.  They work in a way that we will use as a new model moving forward. They deliver real features in less time than it takes most organizations to hold a meeting or provide an estimate.  We've delivered a whole host of capabilities we didn't list at the outset, as well as the original scope. Their core ability is breaking large, risky problems into tiny chunks and learning through direct customer feedback instead of meeting-room theories.

When working with 3wks, was there anything unique or special about them compared to other companies that you may have worked with?

Again, they really do deliver real features in less time than it takes the average organization to hold a meeting or provide an estimate. They always succeed.

And looking back on the project is there any area that 3wks could improve or you would do differently?

No, they were just great overall.

BACKGROUND

Please provide me a brief description of what your company does for background information?

We’re an internal start-up, meaning I operate this small company that sells food to restaurants, but we wanted to start building technology for restaurants. We decided that this was something that we could bring value to.

Where are you located in terms of location, industry, and size?

I’m located in the United Kingdom. The company’s net valuation is approximately £200 million with nearly 800 employees, but the internal start-up is myself at the moment.

What’s your position within the company?

I used to be the Chief Information Officer in the IT division of the main company. I’m now the Chief Executive Officer of this organization.

OPPORTUNITY / CHALLENGE

What was your company’s business goals or reasons for taking on a project with 3wks?

The primary objective was building tech solutions of various sorts.

To build tech solutions for whom, specifically?

The restaurant industry, primarily.

SOLUTION

How did you select 3wks? What was the hiring process?

I was in search of Google App engine development partners. I visited the Google website to find a list of potential partners. Because 3wks shared my belief regarding lean methodology, I was inclined to select them. The majority of the other listed organizations on the Google site were not specifically app engine partners, but rather Google app partners. For instance, many focused entirely on Gmail installation and conversion. I sought developers, in particular.

Can you just tell me more about the scope of the project? Did it include design, development and support?

The project scope encompassed everything. The development models are lean, which requires learning the necessary skills and understanding the separate business components on the fly. You’re coming out and speaking and validating those things. They’ve been helping in Sydney to find customers to use solutions as well as do our work as well.

What parts of the project did you in house yourself or did you just start everything with them?

Everything was formulated with them, without access to the resources of the main company. Because of that, our first step was the identification of candidate market segments. After which point, I was going to need to find and construct a team, which would have taken serious time investment. That’s essentially why I sought an extraneous company to work alongside with.

What technologies were used for the project by 3wks?

They used Java on the Google App engine. On the database side, you don’t really have too many choices with the Google App engine. Google’s got its own priority database on the platform. We had to use that. I knew it was going to happen.

Can you just give me a sense of the size of the initiative in terms of dollars?

I really can’t recall the specific monetary value, but you can imagine as a start-up, the financial situation was relatively tight.

That’s not a problem. And how recent did you finish the project with them?

We established a commitment to build a minimum viable product, the minimal item we need to build so we can conduct preliminary marketplace testing. That was a part of the 3wks growth process. It’s also how we’re discovering what the customers need. At that point we return to the drawing board in order to make necessary modifications and adaptations to the concept. It really isn’t a fixed project with strict milestones linked to static progress reports.

RESULTS & FEEDBACK

So far, I know you said it’s not completely fixed, but what kind of results have you seen? Like specific improvements in effectiveness or efficiency?

Everything is fine when it comes to the efficiency and technology like the Google App engine, which eliminates many headaches that we might normally face when building from scratch. We’re focused on the business results rather than the technical results. Each time we move to another iteration, we’re required to learn something new in order to meet the customer needs.

How did you 3wks perform overall? Did you feel like they covered the agreed scope and the deliverables?

They’ve done very good work so far. They’re also extremely adaptable. That’s the start of this work. It’s not really akin to a standard project where you can evaluate the steps clearly, establish a reasonable scope and begin to build. It’s far more complicated than that. This project isn’t linear. We can’t know what the restaurant industry wants without exposing them to our designs and observing their market reactions. There is really no other way to know whether or not they would be willing to pay for our solutions.

Is there anything special or unique about 3wks that you observed in comparison to other companies you may have worked with in the past?

They fundamentally believe in the lean methodologies, whereas many others are willing to work with you so long as they can continue adding billable hours. Those firms are not truly concerned with the feasibility or success of your business model.

Looking back on what’s happened so far within the project, do you feel like there’s anything that they could have improved on?

We probably need to stay more focused on identifying our unknowns regarding the business model. We ought to invest more time in reflecting upon our internal and external constraints. We’re only eight weeks into the project, which is also essentially a company, so I’m still very much in a haze of discombobulation.

Thank you and do have any further comments regarding 3wks performance?

They’re really focused on designing what you’re trying to focus. They familiarize themselves with the business and then figure out what needs to be done, whether it is technological or something else. Once they’ve identified a potential solution, they begin their diligent work. I appreciate that kind of tenacity and willingness to explore numerous paths.