• Cathal Doyle

Anaplan Data Integration: Where do I start?

Data Integration can cover any manner of sins, but it is undoubtedly on the lips of every customer we meet when enveloping requirements. We encourage the discussion to happen early and often in the discovery cycle so that we, as the Business Partner, can understand what we are in for just as much as the client. And, I would always encourage the ‘keep it simple’ approach. By doing so I’m not saying you cannot automate, customise and enrich your data, but suggesting that you get your base right before you head down the path of most resistance.

As we know, data forms one quarter of the four cornerstones of The Anaplan Way which provides the foundations for the implementation of Anaplan, and it is expected that data will encompass almost 30-35% of an overall implementations timeline. It is the one area of a project that can stop you in your tracks before you even get going and that is why for me, data is the biggest risk to a successful land. Users are not sympathetic to looking at inaccurate or missing data; in fact, the opposite is true and it can be incorrectly perceived as the fault of the application rather than shoddy data sources. Hence, it is critical that there is adequate pre-planning, resource and foresight to ensure a structured approach to Data Integration from the outset.

We now have a wealth of options available to us for getting data in & out of Anaplan, whether that’s by manual import/export within the platform itself, fully automated via Anaplan Connect & API or using third-party connectors like Hyperconnect (an Informatica connector), Boomi, Mulesoft, Snap Logic, Tableau and Docusign. We’ve seen a real shift in the market over recent times as fast-growing companies search for truly integrated systems that support much faster, cleaner business processes and hence Anaplan’s focus on this area of late. However, my recommendation is still to start small, focus on data quality and automate later. The fact that data and metadata is not automated will not stop your project, but poor data will.

A lot of the heavy lifting that was once done in legacy ETL tools, can now be done locally to Anaplan using our Data Hub concept. A Data Hub is a separate model that holds a company’s key data and metadata. Data can then be shared with all other models, making expands easier to implement and ensuring data integrity across models. By starting small, ensuring your data quality is high you can then automate the process of data transfer to the Anaplan Data Hub from a Data Warehouse or upstream ERP, CRM, HRS via any of the available connectors. The key is to find trust in the data before working on the automation. The most encouraging milestone for a business user is to see data that they recognise which leads to the “ok – this actually works” moment which is critical to buy in. Remember, Anaplan should be a single version of the truth, so we need to rely on the integrity of our data.

As companies organically grow, so do their internal systems which will typically be dealt with in silos, one for managing the customer relationship (CRM), one for managing the workforce (HRS), one for accounting (ERP), one for project management and the list continues. Over time, and due to siloed operations, the metadata in these systems will become disconnected, for example something as simple as a customer name might be “Anaplan” in one system and “Anaplan Ltd” in another. On face value, it doesn’t seem a massive problem but companies continue to struggle with marrying their data together which leads to misaligned plans. Of course, as a planning Platform, often, we are required to bring these items together to give a complete view of the business and to be able to analyse and plan ahead. Hence the Data Hub concept, where the time is not necessarily taken in automating the transfer of data but rather in cleansing and enriching the data from upstream sources into a usable and trustworthy manner.

The ‘keep it simple’ approach is not proposed to scare you off, rather to prompt discussion on data early and often so that we can take appropriate measures to ensure we have a successful land. A healthy dose of honesty is required on all sides to ensure a smooth data transition, and more often than not, the automation is the easier side of the discussion.

269 views0 comments