The CRM Data Cleanup Mini-Guide

The most common mistake people make when cleaning a CRM is jumping straight to deleting records. Do that without a clear picture of what you have and you will delete something you needed. Spend thirty minutes running a proper audit first. In HubSpot, go to Contacts and export your full contact list to a spreadsheet. Include these columns at minimum: contact name, email address, company, lifecycle stage, last activity date, lead status, and the date the contact was created. This spreadsheet becomes your working document for the entire session. Once you have the export, look for four things before you do anything else. How many contacts have no email address? How many have not had any activity in the last twelve months? How many have a lifecycle stage of 'Unknown' or blank? And how many share the same email domain but appear as separate companies? These four numbers tell you how much work you have, and where to start.

Duplicates in a CRM are not just a tidiness problem. They mean the same person gets emailed twice, their activity is split across two records so you cannot see the full picture, and your contact count looks bigger than it is. A CRM with 4,000 contacts that is 20 percent duplicates is really a CRM with 3,200 contacts. HubSpot has a built-in duplicate management tool. Go to Contacts, click Actions, then Manage Duplicates. HubSpot will show you pairs of records it thinks are the same person. Review each pair and use Merge to combine them. The primary record keeps its properties, and any additional data from the secondary record gets pulled in. Work through this list first before touching anything else. If HubSpot's tool does not catch everything, run a filter in your spreadsheet. Sort by email address and look for near-matches where someone has used two different email addresses. A person who appears as [email protected] and [email protected] needs to be merged manually. Flag these in your spreadsheet, then go back to HubSpot and merge them one by one. It is slower, but it is accurate.

A contact with no lifecycle stage, no lead status, and no company name is essentially invisible to your reporting. You cannot segment them, you cannot filter them, and you cannot use them to build a meaningful view of your pipeline. Fixing missing properties is the part of CRM cleanup that pays back the most. Start with lifecycle stage. Filter your contacts in HubSpot to show everyone with a blank lifecycle stage. Based on what you know about how they came in, assign them a stage. Anyone who opted in to a lead magnet is a Lead. Anyone you have had a sales conversation with is a Marketing Qualified Lead at minimum. Anyone who has bought something is a Customer. You do not need to be perfect here. Getting everyone into a rough stage is far better than leaving them blank. Next, tackle missing company names. Filter for contacts with no associated company. For the ones you recognise, fill in the company name manually. For the ones you do not recognise, check the email domain. Anyone with a company domain email (not Gmail or Outlook) probably works at a company you can identify. Add it. Then look at your custom properties. If you have properties you created but stopped using, either delete them or decide to fill them in. Unused properties cluttering the record make it harder to see the data that matters.

A contact who has not opened a single email in eighteen months and has never visited your website is not a lead. They are historical data. Keeping them in your active marketing lists inflates your contact count, hurts your email deliverability, and muddies your reporting. You do not delete them. You segment them out. In HubSpot, create a new list called 'Inactive 18 Months'. Set the filter to: last email open more than 18 months ago, AND last form submission more than 18 months ago, AND last page view more than 18 months ago. Anyone who meets all three criteria goes into that list. Then build one more list called 'Invalid Email' for any contacts where the email hard-bounced or is marked as invalid. Once those lists exist, remove those contacts from your active marketing lists. They stay in your CRM as records but they are no longer cluttering your live segments. If you want to try re-engagement before writing them off, send a single 'still want to hear from us?' email to the Inactive list. Anyone who clicks stays active. Anyone who does not gets added to an archived list and excluded from all future sends. This one step alone can improve your email open rates noticeably because you are sending to people who read.

Once you have removed duplicates, fixed properties, and segmented the inactive contacts, you will probably still have gaps. Job titles missing. Company sizes unknown. Industries not tagged. Filling these manually for hundreds of contacts is not realistic. This is where AI enrichment tools come in. The simplest approach for most small to mid-size contact lists: export your contacts to a spreadsheet, then open ChatGPT or Claude and paste in a batch of rows. Ask it to look at the name, email domain, and company name and suggest the most likely job title, industry, and company size. It will not be right every time, but it will be right often enough to be worth doing. Import the suggestions back into HubSpot as a bulk update, then spot-check a random sample to confirm accuracy. For larger lists or ongoing enrichment, tools like Clay, Clearbit (now part of HubSpot's premium tier), or Apollo can automate this. Clay in particular is worth knowing: you feed it a list of names and companies, and it pulls job title, LinkedIn URL, company size, funding stage, and more from public sources. The output is not perfect, but it is dramatically better than a blank field, and it costs a fraction of the time manual research would take. Enriched records let you segment with precision, which means the emails you send are relevant, which means people respond.

After the cleanup, take thirty minutes to rebuild your core segments as saved lists in HubSpot. These are the segments you will use. Think of them as the views you want to be able to pull up any time without having to filter from scratch. The five lists that most small business owners and consultants need: active leads (lifecycle stage = Lead, last activity within 90 days), warm prospects (lifecycle stage = MQL, no deal created yet), customers (lifecycle stage = Customer), past customers (closed lost or churned), and warm re-engagement (inactive for 6 to 18 months, not yet fully gone). Create these as active lists in HubSpot so they update automatically as new contacts come in. Once the lists exist, add a note to each one about what action belongs to each group. Warm prospects get a follow-up sequence. Past customers get a check-in email once a quarter. Active leads get added to your nurture sequence if they are not already in it. The CRM cleanup is only half the job. The other half is making sure the clean data flows into your marketing and sales activity.

The reason CRMs become neglected in the first place is that there is no routine for keeping them current. The data that is clean today will degrade over time. People change jobs, email addresses go inactive, companies get acquired. Without a maintenance habit, you are just scheduling your next cleanup session. Set a recurring task in HubSpot or your calendar for the first Monday of every month. The task takes fifteen minutes. Open the duplicate management tool and merge anything new. Check for any contacts imported in the last 30 days that are missing a lifecycle stage and fix them. Run a quick filter for any new hard bounces and move those to your invalid email list. That is the full routine. Fifteen minutes, first Monday of the month, every month. Also set up a simple automation: any contact who submits a form on your website or books a call should automatically get their lifecycle stage updated and a task created for follow-up. This means new contacts arrive in your CRM already tagged, which is much easier than cleaning up tagging after the fact. HubSpot's workflows make this a ten-minute setup. Do it once and it runs itself.