Why you shouldn’t cut corners in ecommerce

In this movie the Egyptian Queen Cleopatra bets against the Roman Emperor, Julius Caesar, that her people are still great and that she’ll build him a palace in Egypt in a remarkably short time - just three months. This important task falls to architect Numernabis (eng. ver. Edifis) whose previous works were simply not great or even alright but he was the only one available. If he succeeds he’ll be covered in gold, if he fails he’ll be thrown to the crocodiles (hands up who hasn’t met that kind of project in IT before). Since the mission is impossible to complete with normal solutions, his only option is to seek help from Gaul druid Panoramix (eng. ver. Getafix) who can brew a magic potion giving supernatural strength.

You may wonder, why mention all this?

Because in real life there are no such wonderful ways to overcome unrealistic deadlines. Hurrying up a project launch will certainly lead to minor or major problems. Cutting corners in ecommerce, although it may be tempting, in the long term will require additional time and work, not to mention money, since rushed results often are far from expectations.

The plan

One way of saving time is skimping on planning, where for example some parts are only thought in general or placed temporarily. Without a correct and thorough design it’s really easy to misplace something or overlook mistakes. Of course it happens because of faulty implementation and insufficient tests, but still, with better preparation it is much easier to avoid. On some of the audited stores it showed up in things like:

• unintuitive placement of products, categories and sub-categories (product tree structure with different branches for similar or same products),
• switches for language or currency with empty or mixed options,
• incorrect images or thumbnails
• “chaotic” product descriptions
• parts of pages covered by popups, hindering user experience
• alignment of page elements on mobile,
• some possible user actions (like going back on certain forms) were not predicted and resulted in incorrect page behavior or errors.

The meaning of the good analysis

Another way is not doing a proper project analysis for possible issues, weak points or bottlenecks, either from the Devs or QA perspective. Even if you plan on using market solutions, everything on your current platform is working alright and changes may look simple, there could be unforeseen obstacles which will require additional time and work.

One of the examples that comes to mind was adding a blog on Wordpress to a Magento store: the basic integration went smoothly but problems occurred with the user menu and cart and the way the cache was stored.

Another time it was with checking Magento upgrades with implemented payment gates and widgets, and if there would be issues with their interactions. Some unlisted behavior was found which required a workaround. Luckily it happened in a staging environment and not in production.

“It works” vs. “It works always”

QA verification often also falls victim to short deadlines. The preparation of thorough tests scenarios with a broad range of values and their later execution might take a lot of time. The importance of testing is often misunderstood. For example spending less time on tests which show no issues doesn’t prove there aren’t any problems. Testing some parts without finding any problems may lead to thinking that other parts will also be alright. That makes it look like the best place to make some time savings. The results of thinking that way may vary but most times a simple “It works” is not be the same as “It works always”, and when the issues finally come up they usually make everyone's day much harder. The correct amount of time for QA (and for possible fixes) should always be incorporated into the project time schedule and never skipped.

Finally there is optimization

Many times it’s overlooked to prioritize the product launch but it shouldn’t be. Even when some solutions are good enough at project launch, they often quickly show their limitations and sometimes even render pages unusable. Many times overcoming these new problems takes much more time than when optimization is planned from the beginning. Some good practises are:

• only using correct and needed CSSs and JavaScripts,
• proper next-generation media formats
• sufficient sizes to save cellular data and speed up loading
• lazysizes for lazy-load.

Hard work is a magic

It would be nice if there were magical solutions or heroes capable of astonishing deeds to help to advance projects but reality is brutal: the time and work required cannot be reduced with wishful thinking. There are ways for apparent workarounds but they have their implications and more than often come at a price. Everyone wants their business to be treated seriously and be taken care of the best. And that is reasonable. However that requires both effort and realistic time frames for ecommerce projects to meet expectations.