The test lifecycle is a curious animal involving several steps such as ideation, development, and quality assurance. But one thing is certain; everyone benefits when a new test flies through the coding phase. This might feel outside the marketer’s control, as they must rely on IT support for the follow through, but there are a number of ways to reduce development time and overall efficiency of the test creation life cycle. Organized developers and savvy marketers can work together to speed up the development time of a test.
I came up with eight ways to streamline your test development process from the outset. Print these out, pin them up, and watch the timeframe of your test lifecycle shrink down a size or two.
1. Keep your Workspace Organized
This is one of the most important and easiest ways to boost campaign development efficiency. When your workspace is organized, all objects (pages, actions, campaigns, scripts, variants, etc.) are easier to find and navigate, existing functionality is easier to understand, and reusing existing objects and code becomes effortless. Come up with a consistent naming convention for all objects in your workspace, write clear descriptions when creating new objects, and regularly audit your workspace to remove old and unused objects.
2. Be Aware of How your Code Will Function in Earlier Browser Versions
When adding any new code to your testing workspace, keep in mind that what you are writing will be running on all included browsers. Make sure you are writing code that is compatible with all relevant browsers and their respective versions. It is not uncommon to code up a large variant only to discover that an essential piece of the code does not function properly on Internet Explorer 8 and now has to be revamped. Save time by discussing coding efforts in older browsers before development begins and coding for all browsers up front.
3. Follow General Coding Best Practices
Many developers do not follow coding best practices in their testing workspaces and end up spending too much time trying to understand their old code or another developer’s code. If you have multiple developers sharing code in the same workspace, following coding best practices will allow each developer to interpret the other’s code more easily, further reducing the time it takes to understand existing functionality.
There is a lot of time to be gained by adhering to some very easy guidelines. Descriptive commenting, consistent indentation, limiting line length, and useful variable names can go a long way in saving time now and for future tests.
Be on the look out! Writing duplicate code can be another hidden time sink. Store duplicate code in functions defined in global scripts instead, and then call those functions where needed.
Lastly, group similar code together in the same script. For instance, keep all code related to calculating and collecting custom user attributes (gender, favorite product type, etc.) in one function in a global script, then call that function anytime you need to update the user’s attribute values and send them to a test.
4. Watch out for Environment Differences
If you have to do your development work in a pre-production environment (such as uat.domain.com or staging.domain.com), familiarize yourself with all of the differences between the pre-production environment and the live environment before beginning development so that you avoid spending development time fixing bugs. The most optimal way to code a test is to use your testing platform’s staging/sandbox environment if it has one. This will allow you to code against your live site’s production code.
5. Don’t Allow the Default Production Site to Change
Changes to the default site are a problem whether they happen during development or while the test is already live. Always know what changes are being pushed to the production site and when they’re arriving to avoid the possibility of needing to go back and change test code that has already been written.
6. Make Good Use of Third Party Integrations
Some campaign metrics are best tracked using your existing analytics platform. Most major analytics packages can be integrated by sending aggregated data about which population of users saw which experiences of a test, allowing you to analyze and compare these populations in that analytics package. Most metrics are best coded in your testing platform itself, but spare yourself hours of development work coding trivial metrics that will yield the same information as your analytics package. In addition, because of the nature of how these metrics work, some metrics such as bounce rate are best monitored in an analytics platform.
7. Flush Out All Possible Campaign Use Cases
When developing a test, all possible page experiences and scenarios should be accounted for to avoid bugs and to collect clean data. You don’t want to code three quarters of a test only to realize that everything needs to be re-written to account for a special experience on the search results page that only applies to certain search terms. Before development, get all relevant departments in your organization briefed on the campaign’s experiences and metrics so they can point out any use cases you may not know about or may have forgotten to consider.
8. Attend Maxymiser Academy
Attending Maxymiser’s business and development training is by far the best way to become the most efficient Maxymiser developer. Not only is everything above covered in detail, but training will also teach you the most efficient way to set up every type of campaign while even showing you many features and tricks that can’t be discovered anywhere else. You can sign up for training and get course details on the Maxymiser Academy page.
Whether Maxymiser is your testing solution or not, these tips will help speed up the development of your test campaigns. For more on streamlining the test set up process, check out our post on how to keep your campaign concepts simple for strong metrics and fast results.