Data

Best Practice Guide to Sales and Marketing Data

By Andrew Nicholson · Published 26 April 2025 · 7 min read

When having a sales conversation with a new prospect, some of our first questions are always about the quality of their contact data: where it came from, when it was last cleansed, how it's segmented, and where it actually lives. It's surprising how often even large agencies and brands have no good answer. Data, alongside content, is the bedrock of any marketing automation strategy - bad contact data means bad targeting, and bad targeting means even great marketing falls on deaf ears.

Key findings

  • Bad data is estimated to cost sales and marketing teams roughly 550 hours and up to $32,000 per rep per year (DiscoverOrg).
  • Good data has six qualities: clean, fresh, tidy, rich, unified, and legally compliant.
  • GDPR doesn't require opted-in data - legitimate interest is a valid legal basis - but the data subject still owns their own data, and you must make it easy for them to access, update, or delete it.

1. Good Data Is Clean Data

The fastest, cheapest hygiene exercise is running your contact data through a third-party validation tool like Neverbounce, which flags each record as:

  • Valid - a real, currently-accepting address. Safe to send.
  • Invalid - doesn't exist or isn't accepting mail. Don't send - it will bounce.
  • Disposable - a temporary address (Mailinator, Guerrilla Mail, etc.). Don't send.
  • Accept All / Catch-all - the domain accepts everything, so there's no definitive way to tell if it's valid. Bought data disproportionately falls into this category - we don't recommend using it.
  • Unknown - undeterminable due to a non-responsive domain or server. We don't recommend using these either.

2. Good Data Is Fresh Data

Look honestly at your open and engagement rates. What percentage of your contacts haven't opened anything in 6 months? 12 months? Ever? Repeatedly emailing disengaged recipients damages your domain reputation and therefore your deliverability across the board - email providers actively use past performance to decide whether you land in the primary inbox, promotions, or spam. Our advice: accept what the data is telling you, suppress unengaged contacts, and if it feels too final, send one last re-engagement email with a clear opt-back-in action - just don't be surprised if very few take it.

3. Good Data Is Tidy Data

By tidy, we mean a proper, relational CRM that recognises contacts, organisations, and deals as connected entities in one central location - not a spreadsheet (Excel is not a CRM). Salesforce is powerful at enterprise scale but often overkill - and expensive - for smaller teams. HubSpot's free CRM tier is easy to start with but gets costly quickly as your needs grow. Pipedrive tends to hit the middle ground well: powerful, reasonably priced, and with solid workflow automation on its Pro tier - which is part of why we use it ourselves and are a Pipedrive partner.

4. Good Data Is Rich Data

Forename, surname, and email address alone isn't enough. Rich data lets you target and segment far more effectively:

  1. Comprehensive CRM data - tools like Clearbit or FullContact enrich an email address with job titles, social profiles, and more.
  2. Behavioural data - what pages prospects view, what they download, what they click. A lead browsing your pricing page is far more likely to convert than one who's only seen your homepage.
  3. Lead score data - once you can track online behaviour, you can score it, automating follow-up sequences, CRM updates, and sales alerts off the back of score thresholds.

5. Good Data Is Unified Data

Ideally, your CRM is the single source of truth, consolidating data from every inbound channel into one rich record. In reality, most CRMs struggle to "talk" to every other system perfectly. Three practical options: accept the limitations and focus on the 20% of data that drives 80% of the benefit; use ETL tools like Zapier to unify formats as data passes between systems; or use a business intelligence tool like Tableau to clean and unify data before loading it back into your CRM. Duplicate records carry their own cost too - commonly estimated at around $1 to prevent, $10 to correct, and $100 to leave untreated (SiriusDecisions).

6. Good Data Is Legally Compliant Data

GDPR is as relevant today as it was in 2018. A few common misconceptions worth clearing up:

  • Your data doesn't have to be opted in. You can legitimately use third-party or bought data, claiming legitimate interest as your basis for processing - just make sure your data processes and policy are clearly documented. (We still don't recommend rushing out to buy a list from a disreputable broker.)
  • It's not your data. It belongs to the data subject - you're only looking after it. If they want to update, delete, or export it, you must allow that, and make it easy.
  • If you don't need it, you can't have it. Only hold data that's relevant to the service you actually provide them.
  • It's not yours forever. You can only hold data for a legitimate period - tied to the nature of what you sold them, clearly documented.
  • It's not global. Moving data across borders, even just in transit, can break GDPR if the destination handles data security differently - audit any third-party data processor and document where data is actually processed.

None of the above is legal advice - for the full detail, the ICO's guide to data protection is the right next step. Get in touch if you have questions about your own sales and marketing data.

Frequently Asked Questions

What are the six qualities of good sales and marketing data?

Good data is clean (verified and bounce-free), fresh (regularly updated, with disengaged contacts removed), tidy (structured in a proper CRM, not a spreadsheet), rich (enriched with behavioural and firmographic detail), unified (consolidated into a single source of truth), and legally compliant (GDPR-aligned).

Does GDPR mean all marketing data must be opted in?

No. Businesses can lawfully use third-party or purchased data under the legitimate interest basis, provided data processes and policy are clearly defined and documented. It is not automatically illegal to use non-opted-in data, but the legal basis and documentation must be in place.

Andrew Nicholson

Andrew Nicholson

Co-founder & CEO at Kulea. Guest lecturer in digital psychology and behavioural economics at the University of Buckingham.

Ready to fix your data foundations?

Book a demo