We can talk about innovation and the importance of staying up to date as a company, but sometimes it’s easier said than done. At the Agile Tour recently held in Quebec City, my colleague Francis Côté and I explained how creating startups at Spektrum has allowed the agency to become more technologically agile.
Quit your job, stay in the office
First, our interlocutors are always surpris to learn that at Spektrum, it is possible to leave your job to develop a product, while being support by the company. With the idea that human aspirations are constantly evolving, the agency’s co-founders did not want to lock their employees in boxes .
This is how Signsquid , Snipcart and Userboa t were born, a membership management software that I slovenia telephone number data work on over the last year. These initiatives are at the origin of what has become the “intrapreneurship” component of our Apollo13 startup program.
No experimentation on the customer’s back
By using emerging practices, each project has enhanc Spektrum’s technological know-how. Beyond the financial aspect, it is the best return on investment we could have hop for. The and abroad, using our experience, we provide team has even come to change its approach to developing projects in service mode. The nuance is important:
in an agency context, experimentation with the aim of innovating should not be done on a client budget. It is unhealthy and disrespectful. Generally, a client pays the agency for its expertise, in short, to do what it knows best (unless it is in R&D). Developing products in-house and being our own client, with our own budget and schule, is much more conducive to evaluating and testing new methodologies.
Techno impact: let’s talk geek
Warning: with this passage, I will probably put our aspiring geek in the juice with his popularization chronicles! Our product approach allowed us to pivot in a very short time from the traditional world of the multi-page application and the showcase site support by a CMS and dedicated servers, towards a single-page webconnect by API directlyserv in the cloud. We were able to integrate technologies such as Vue.js and serverless architecture into our daily lives .
Also, we increas our knowldge of Microsoft Azure and Amazon Web Services, services with inestimable value because of the quality they allow to achieve and their minimal operating cost.
Take the means
How did we realize our desire to develop products in-house? First, you have to be able to dedicate yourself completely to the project . For Userboat, my schedule was freed from all the agency’s clients. Several mandates were postponed or put on hold in order to make the operation possible.
We also convinced a client to jump into the adventure with us, since our idea met his exact need. For a much lower cost than what would have been incurred by creating a custom system, he was able to equip himself with a platform adapted to his needs, in addition to obtaining access to customer service for life. In the event that Userboat closed its doors, Spektrum would then be responsible for maintenance.
The rule of the four Rs
We can summarize our decision-making process with the four Rs of good tech choices : inventory, rationalize, burnish, and spread.
This system frames the technology update process to ensure that we make the right choices in different contexts.
1) List: know your options
It all starts with clearly defining our needs. Since we were in an MVP ( Minimum Viable Product ) context, that is, the smallest thing that can be developed that will bring value to the customer, we had to deliver the platform as quickly as possible while minimizing costs. In such a context, several uncertainties arose, such as:
- Implementing features similar to those offered by a big player like Mailchimp;
- A data schema that varies greatly depending on the customer’s reality (like Salesforce);
- The ability to deliver significant performance regardless of the client, meaning that a single client can send email campaigns to 1 million people if they want.
To develop such a tool, we needed a malleable solution that we could use easily. And since Mailchimp cannot be replaced with the snap of a finger, we needed a reliable strategy to meet different challenges such as managing email campaigns, sending without having to manage the servers ourselves, bounces , unsubscriptions and spam notifications. notifications , in addition to integrating the solution into the platform’s dashboard in a smooth way.
There were a few options to meet these requirements. Obviously, we couldn’t develop everything from A to Z. During our research, we target two interesting tools: Sendy and MoonMail .
#geektalk alert here too. The choice of database management system is always a source of argument. The eternal debate between storing structur (SQL) and unstructur (NoSQL) data arouses passions and unfortunately. Making the wrong choice can have serious consequences. Our ne was not standard, i.e. a hybrid between structured and unstructur. So, what to do?
Two databases for the same application? Store everything in relational form and live with the consequences? The implementation variants were multiple and the potential tools were even more so (SQL Server, PostgreSQL, RavenDB or MongoDb).
In short, listing means knowing our options. To do this, we need to stay on the lookout for new developments and understand their add value. Ideally, listing is a continuous action . That way, when a new ne arises, we have in mind a catalog of tools within our reach.