Salesforce Data Cleaning Tools: Fix Duplicates, Standardize, and More

Everyone is talking about AI. They want to embed it in the organization somehow – no matter how. Have you already been asked to “add some AI” to your org? Here’s the thing: AI won’t give you good answers if your data is poor quality, and most organizations still struggle with Salesforce data quality.

Unfortunately, many companies neglect data quality, yet they still expect accurate reports, high-quality leads, and seamless integrations. As a seasoned admin, you know that you need accuracy to build trustworthy dashboards, clear boundaries to capture quality leads, and uniqueness for integrations to work effectively.

I currently work for startups, and guess what? Investing in data quality is often low on their priority list. So, in this article, I’ll walk you through a few lower-budget tools I’ve been evaluating. While they may not be as robust as full enterprise-grade solutions, they’re strong enough to get you to safe shore.

What Does Salesforce Offer for Data Quality?

In an ideal world, you wouldn’t need any third-party tools. You’d have perfectly defined a precise set of required fields, validation rules, and duplicate rules that would prevent a single incomplete or duplicate lead from entering the system.

Are you the only person in that org? Is that a dev-org by any chance?

The truth is that messy data always finds its way in. And once it’s in, out-of-the-box Salesforce doesn’t offer much help. The Duplicate Record Sets report only gets you so far, and deduping still takes a lot of manual effort.

Key Problems These Tools Should Solve

Several tools claim to solve all your problems, but which issues should we be asking them to solve? That’s a key question, and we need to consider it when evaluating these tools. Our demands are clear:

For handling duplicates:

• We will ask them to find duplicates.
• We will ask them to let us define custom, complex matching criteria to find those duplicates.
• We will ask them to get even more sophisticated rules when deciding which record should win – ideally using advanced, customizable rules.
• We will ask them to merge those duplicates.
• We will ask them to properlyreparent related records after merging.

For handling normalization:

• We will ask them to standardize fields like phone numbers and state codes.
• We will ask them to apply consistent capitalization.

And, without getting too greedy, we will ask for a free trial.

Of course, there’s more we can ask for, like enhanced data loading features. Still, I imagine you’re already proficient with tools like the quick-and-dirty Salesforce Inspector, the good old Workbench, the faithful Data Loader, or my new favorite, Jetstream. I only see value in a data loader when it has capabilities like upserting leads or contacts in a single load. That’s a huge time saver!

If you’re after a solid overview of data management tools across the Salesforce ecosystem, I recommend checking out Lucy Mazalon’s and Mehmet Orun’s articles on Salesforce Ben. They do a great job mapping the landscape and explaining how these tools fit into your overall strategy.

A Deep Dive into Three Data Cleaning Tools

So, I’ve been doing some tool window shopping, and I found some worth mentioning. Some of them you might already be familiar with, but here you’ll find an honest opinion, and I’ll always be trying to answer the question: do they do the job?

Insycle – The Full Data Hygiene Hub

When I first bumped into Insycle, I was impressed with the number of CRM connections they provide. It can connect not only with Salesforce but with many other platforms, so I see a great use case for companies that have a diverse tech stack, like Salesforce for sales, HubSpot for marketing, Zendesk for support, and so on.

After trying it out, I found that you log in with your credentials, and it immediately generates an assessment. This analysis will highlight duplicates, poor formatting, or inconsistent, incomplete data, and you can tailor it for future runs. The combination of this assessment and some pre-built templates gives you a great place to start if you’re new or don’t want to invest too much time playing with it.

The deduplication tool is quite comprehensive. It has a few pre-defined functionalities, like what to ignore and what to match within a field (e.g. last n characters), but I found the comparison rules a bit limited (only exact and similar).

You can, of course, apply some logic to define the master record, but nothing cross-object (e.g. the latest related closed-won opportunity). After the duplicate analysis is done, you can see the results in a CSV file, but to do so, it will email you the file, which I found a bit inconvenient. From there, you can perform the merger.

It also includes a transform tool for basic string manipulation, such as splitting data into fields or extracting names from email addresses.

Insycle also has powerful import capabilities. For example, it can match an account to a contact upon insert with complex criteria. Additionally, you can restrict your imports with validations to prevent user error. It offers interesting nuances on actions beyond insert and update, like append or overwrite. However, it does not include the most desired capability for Salesforce users (which I see might not be a need for other CRMs): the ability to upsert contacts or leads in a single load. You can run the import on one or the other, but not both simultaneously.

Other tools under its belt include the grid (which works like an inline editor), cleanse data, and grouping, which can help you with harmonizing picklist values, to name a use case.

Insycle offers a 14-day free trial where you can merge up to 500 records. Regarding pricing, their subscription model is based on the number of records you connect with each CRM. It might make sense for a small book of accounts and a curated list of prospects, but it could quickly and unexpectedly increase the licensing cost, considering it also counts against opportunities and users. Overall, I find it a comprehensive tool, particularly handy for data hygiene issues that exceed Salesforce’s ecosystem.

DemandTools Elements – Cloud-Based Deduplication Power-Up

I’ve always been a big advocate for DemandTools. It’s a powerful, easy-to-use, and dependable tool. But if you’re on a tight budget, the full version will very likely be out of your league. Luckily, they have a lower-tier offering called DemandTools Elements, which I’ve been testing to see how much I would miss the big one. Here’s my take: while the traditional DemandTools platform is a robust, locally installed application, DemandTools Elements provides a simplified, cloud-based solution focused on core data quality tasks, making it accessible for a wider range of users.

If you’re struggling with data manipulation, duplicates, and normalization, it does it all. It has pretty much the same features as its older sibling, with the big misses being the Lead-to-Contact match and convert feature and the ability to match data with external resources through files (which is useful for migrations, for example).

You can define the same complexity for deduplication as DemandTools V, including fuzzy matches, different kinds of street and phone matches, field logic upon merge, winner rules, and more. It also includes what was once called People Import, which handles the much-needed upsert of Leads and Contacts. You can do this under the Import tab, and it works wonderfully.

The subscription is based on the number of licenses in your org, and it’s about 60% cheaper than its robust older brother. You can also get a 14-day free test drive.

DataGroomr – Taming Duplicates with AI and Machine Learning

Next, I wanted to try something edgy, and I found DataGroomr. It has an excellent range of tools for data manipulation, including deduping, cleansing, and matching. It also introduces the “transfer” concept, which is for creating records in one object based on another. Additionally, the Professional edition includes mass conversion, which is a nice touch.

What I liked the most about this tool was the introduction of AI, specifically machine learning. So, what does this mean and how does it work? For deduplication, you can define “matching models”. In other tools, you would define more or less complex criteria based on some field values, and that’s it. But here, after you build your mix, it runs an analysis and displays a few matches. You then get to choose whether it’s a good match or not based on your judgment, which trains the tool. 

Do that for about 20 records, and that machine learning will be applied to future matches if you choose that model. My data steward past self rejoices with this feature; if you’ve been checking hundreds of matches and always trying to fine-tune the best criteria ever, this was the ultimate feature you were looking for.

Besides that, it introduces the Match Confidence concept, which is a bit more sophisticated than what other tools offer.

For data quality, it offers a feature where you can define different data quality models to do what, for those familiar with Pardot, we would call grading. Even though it’s a nice addition, it’s a bit more complex than the other modules, so I foresee a steep learning curve here, but it’s definitely a must-try.

But not everything is nice and groundbreaking in DataGroomr. It doesn’t have a feature for upserting leads and contacts simultaneously – it’s one or the other, which is a significant miss.

For the rest, I think it’s a great tool, the most innovative so far. However, I believe it lacks a few basic capabilities, but still a strong contender.

It also offers a 14-day trial, and the licensing model is a bit more complicated than the others. It’s records-based (starting at 100K records) plus the number of users of the tool, which I foresee might leave this tool on the pricier side.

Other Things to Consider

• Custom Objects: Do they work with custom objects? The short answer is yes, but for Insycle and DataGroomr, this is considered an additional feature or a higher subscription tier. So, if you need, for example, deduplication on custom objects, you might want to look closely at DemandTools Elements.
• Email Validation: These tools also provide features for email validation (for avoiding email bounces, for example), but if you’re interested in this, be sure to read the fine print, because it might be an added cost.
• Compliance: And of course, these tools are GDPR compliant and uphold security measures, encryption, and such.

At-a-Glance Comparison

Please note: Pricing is subject to change and may vary based on your specific needs and license count.

Summary

And there you have it. I would like to try out a few other tools, like Cloudingo and DQE One. I’ve found them to be good contenders, although their trial paths are a bit more challenging to start. But let me know in the comments if you’d like to know more about them and how they stack up against the others, and I can aim for a part two!

To wrap up, we know that sooner than later, we will face the need to invest in data quality if we want a healthy org, and there are some interesting options to tackle the task in fun and pleasant ways. I think the three tools I mentioned are great in different aspects, and hopefully, this article will help you understand why.