The choice to use a PIM can be daunting. Even after choosing your PIM solution it can be difficult to understand how much time it will take to get your new PIM up and running. As a certified Akeneo implementation Specialist I am frequently asked to estimate work for migrating product data into Akeneo. Far too often implementers ask too few questions before estimating work. Most commonly I’ve seen people asking about data volume and expecting that to be enough information to give an estimate. While information gathering is important for estimation, there are other components of an Akeneo implementation that will have a much larger impact on the project’s complexity.
Having a large count of skus, attributes, attribute options, assets, categories, or associations is one of the easiest things to account for when implementing Akeneo. The total amount of base data actually has a low impact on the level of effort required for an Akeneo implementation. Increasing the volume of these data objects will increase total required processing time linearly. Doubling the total amount of skus in Akeneo will also double the total time to process all of the skus in the system. The main commonality between these 4 data objects is that all of them are strictly defined by Akeneo. As a result the process for adding these objects is frequently automated. Running your automated process for 10 hours instead of 1 hour might be inconvenient, but it doesn’t have long term implications. Typically these objects are only going to be added to Akeneo 1 time and will rarely need to be updated. As a result, the total cost of working with large volumes of data is not felt on a day-to-day basis. You feel the delay once and can move on.
Major Contributors to Complexity, Time and Scope
There are some Akeneo objects that can create long-term issues when using your PIM. These medium impact items include product models, and the number of families, locales, and channels. The commonality between these Akeneo objects is that they act as multipliers on your data catalog. Unlike data volume where you scale difficulty linearly, these medium impact objects scale difficulty exponentially. Adding a second locale doubles the potential data that can be stored in your PIM. If you also add a 2nd channel you’ve again doubled your possible data for a total of four-times more data. Increasing the use of these Akeneo elements can quickly stack up to a much more difficult implementation. What’s more is that these complexities don’t go away. Every time you add an attribute or product data you’ll need to choose the locale and channel. If you’re not careful these medium impact issues could have major implications on the day-to-day of your data management team.
Lastly, there are some high impact elements that fall into 2 categories. First are the PIM-within-the-PIM Akeneo objects: asset families and reference entities. The process of setting up asset families and reference entities is essentially the same as setting up your products and families, except on a smaller scale. Both of these PIM objects have their own attributes with types and options. They also both bring their own unique challenges which contribute to the overall complexity of the implementation. Asset families have naming conventions, product links, and transformations. Reference entities can be included as attributes on other reference entities. The result is that both of these elements are highly unpredictable in the amount of time and effort to complete them. Because these are miniature PIM implementations they should be treated as such.
The second category of high impact components within an Akeneo implementation are connections to outside systems; which applies to both inbound and outbound connections. Akeneo has done an incredible job of maintaining its API. They’ve maintained very robust documentation, which makes connecting to the platform through the API very easy to navigate. Akeneo is platform agnostic, meaning it can connect to virtually any outside system. There are even a few marketplace offerings to integrate with common platforms like Magento and Salesforce. All that being said, even with all of these conditions in place, connecting your Akeneo instance with another platform is the most difficult portion of any implementation. Most likely you will be connecting to a platform that you have never worked with before. On top of that both Akeneo and your target platform are constantly changing versions. There will always be a large amount of mapping and security work when transferring data between platforms as well. If you don’t have an expert resource for guiding you through the connection it is very easy to get stuck at this step.
The Bottom Line
Determining the difficulty of an Akeneo implementation involves much more than simply knowing the volume of data you’ll be handling. Having a poor understanding of what causes Akeneo complexity will hurt your both short and long term. If you have interest in using Akeneo, but don’t feel confident in answering these questions yourself, our experts at Sitation are happy to help.