Not long ago we wrote “Salesforce and data collection apps: a powerful alliance”, highlighting how important information is in the correct functioning of a monitoring and evaluation system. Salesforce is one of the most widely used platforms for handling data in social projects. Here, we want to explore a few alternatives to the typical data entry process. We talk about bulk uploading data, which can save a lot of time and also avoid common human errors in the process.
As we mentioned in the previous post, a bulk upload lets you upload many different records into a database at the same time. To better understand what this is about, think of a typical example in social projects: organization X is training beneficiaries and needs to keep track of assistance. We could take two different paths in order to do this: the first option is to upload each assistance individually, for each one of the beneficiaries enrolled in the program. As you can imagine, when we are dealing with large projects providing lots of training and reaching a huge number of beneficiaries, doing something like this would be extremely time consuming, even virtually impossible; besides, the greater the number of individual uploads to make, the higher the chances of a human error happening. Fortunately, we have bulk upload as a second alternative: we would need to create a spreadsheet in which each row represents a different beneficiary, and each column represents a different time period –days, weeks or months depending on the project’s needs. Once ready, the entire spreadsheet can be uploaded into the system.
Now that we are on the same page as to what bulk uploads are, we can go through a few alternative ways to do them. From our experience, there are two main ways to do bulk uploads in Salesforce: either using an app like APSONA, Dataloader.io etc. or using customized interfaces with Visualforce.
It’s a framework coded within Salesforce that allows developers to create custom user interfaces. In our example, we could create an interface such as the one in the image below, in which different beneficiaries are in rows and the month of training in columns. Since these kinds of interfaces are created ad-hoc according to the users’ needs, the experience is generally much more pleasant than with other methods, as well as faster and more intuitive, especially compared to uploading each month of assistance at a time. Another advantage of Visualforce is that the bulk upload can be performed within Salesforce itself, without having to work with external databases or Excel sheets. This is not an irrelevant advantage, especially if the team in charge of uploading data is not trained and experienced in handling databases and complicated spreadsheets. Unfortunately, Visualforce has a significant disadvantage in the medium to long term: you need developers to create the interface, and if eventually you find adjustments that you would like to make, you would again need developers to work on it. This translates into a significant loss of autonomy in the process, as well as higher costs.
It’s an app that lets you do bulk uploads from CSV files. It has an online version as well as a downloadable desktop version for PC. In order to upload data from dataloader.io to Salesforce, you first need to create a spreadsheet, which needs to be structured in a way that maps the columns to the fields of the Salesforce object involved. Once you have your spreadsheet ready, the app lets you map the columns and the fields in Salesforce. After mapping the variables, you can upload all the records that you need. In our example, the spreadsheet would have all the beneficiaries in the rows, and the months of assistance in the columns, allowing you to input the number of days assisted in each month in the cells. The process would be something like this: people responsible for each training session would keep track of assistance in their own institution/neighborhood on an Excel sheet, which then needs to be sent to the system administrator. This person would then consolidate all the different spreadsheets into one large assistance sheet that can be bulk uploaded into Salesforce. The main advantage of this app is that it’s universally accessible: it has a free version that covers the needs of even very information-demanding projects. The main disadvantage is that it can only be used by someone who has the role of system administrator: because it doesn’t let you restrict the scope of the uploads, anyone other than the admin could potentially intervene the entire system if that person isn’t careful. This can limit the usefulness of the app, overload the administrator with work, and make users excessively dependent on the admin.
This app is conceptually similar to dataloader.io but has some substantial differences. Although it also requires you to build an ad-hoc spreadsheet, APSONA is not limited to CSV format only. It also allows you to copy the content of your Excel spreadsheet and paste it on APSONA’s window directly. You can even copy data from Stata and SPSS. Once you have your data, you can continue editing it, which is useful if you happen to catch any errors in the process of matching variables. Another advantage, in our view the most relevant, is that it allows all different type of users to perform bulk uploads of different Salesforce objects; you can even define which fields within a specific object a user can or can’t manipulate. Going back to our example: people in charge of each training session could create a spreadsheet, then copy all the content into APSONA, map the variables to the fields in Salesforce, and finally do the upload themselves. This way, each member of the team can share responsibility in the data collection process (which is the main disadvantage found in dataloader.io). Unfortunately, APSONA is a paid app, which limits its accessibility, especially for projects with low budgets for M&E.
Having considered different alternatives to the bulk uploads in Salesforce, we can say that there isn’t one obvious choice that stands out. Instead, we see different alternatives that can be useful depending on the case. The project’s characteristics and its M&E needs will help define which of the alternatives is best-suited to help in the data collection process.