HubSpot is one of the services I list on my own work with me page, next to AI workflow automation and AI lead generation. That's because most of the AI work I do ends up in a CRM. A lead comes in, gets researched, gets a reply, gets a follow-up, and the whole thing needs to be recorded somewhere your team trusts. For a lot of businesses, that's HubSpot.

I come at it as a marketer first. I've been doing this since 2003, I've worked with brands like Twitter, Dropbox, IBM and monday.com, and I've reviewed more than 300 tools along the way. So my HubSpot builds start with how you win customers, not with a list of features to switch on.

What I do inside a HubSpot account

Every account is different, but the work usually falls into these jobs.

• New builds: properties, pipelines, lifecycle stages, forms and lists set up around how you sell.
• Rebuilds: untangling an account that grew messy, with duplicate properties and workflows nobody remembers creating.
• Workflows: lead routing, nurture emails, task creation and deal follow-ups that run without anyone remembering to do them.
• AI integration: HubSpot's own AI features plus connections to other tools through Make, Zapier or n8n.
• Training: short sessions so your team uses what's been built instead of reverting to spreadsheets.

The AI settings to check before anything else

Here's the detail most people don't know. In HubSpot, AI is controlled from one settings page: the settings icon, then Account management, then AI. The HubSpot Knowledge Base says Super Admin permissions are required to manage those settings, so if nobody on your team is a Super Admin, nobody can change them.

On that page there are separate switches for giving users access to generative AI tools, for Breeze Assistant, and for what data the AI can use: CRM data, customer conversation data and files. There's also a model training setting. When it's off, HubSpot says your data isn't used to train its AI models and isn't shared with other HubSpot customers, and the AI features still work. I check these with every client before building anything, because the right answer depends on what's in your CRM.

How a HubSpot project runs with me

I keep projects in a fixed order so you see working pieces early.

1. We map how a lead becomes a customer today, including the bits that live in someone's inbox.
2. I review the account, or plan it from scratch, and agree what gets built first.
3. We settle the AI settings and data access with whoever owns customer data.
4. I build the core: properties, pipelines, forms and the first workflows.
5. Your team uses it for two weeks, we fix what's awkward, then we add AI and the next layer of automation.

Where AI helps most in HubSpot

The biggest wins I see are in follow-up and admin. Drafting the first reply to an enquiry, summarising a long email thread before a call, cleaning up contact records, and writing sequences in your own voice. Outside HubSpot, I connect research and enrichment steps through automation tools so new leads arrive with context. My page on AI automation for sales follow-up shows the pattern I use most.

Signs your HubSpot account needs a consultant

You probably need help if your team keeps a separate spreadsheet because they don't trust the CRM, if leads sit untouched because nobody knows whose they are, or if you're using HubSpot as a glorified address book. Another common one: an AI feature got switched on months ago and nobody can say what data it can see. None of these are unusual. They're what happens when a CRM grows without someone owning it.

HubSpot AI or a separate AI tool?

People often ask whether they should use HubSpot's own AI or bring in another tool. My answer depends on where the data lives. If the job runs entirely on CRM records, such as summarising a contact's history or drafting a follow-up, the built-in features are usually the simplest choice because nothing leaves the account. If the job needs outside research or a specialist model, I connect that step through automation and write the result back into HubSpot, so the record stays the single place your team checks.

Who I work with

Entrepreneurs, service businesses and founders, mostly small to mid-sized teams who want HubSpot to run their lead generation and follow-up without hiring a full-time admin. I work over video and in English, for businesses in the UK, the US, Canada, Australia and elsewhere. If you'd rather start with a wider look at your setup, my AI audit for small business covers the whole operation, and every way we could work together is on my work with me page.

My advice before you hire any HubSpot consultant

Ask them how you'll win customers before they ask which hub you bought. A HubSpot build is only as good as the sales process it records, and the best consultants spend the first conversation on that. Then make sure someone on your side is a Super Admin who understands the AI settings, because that decision shouldn't sit with an outsider forever.

Frequently asked questions

Who is a HubSpot consultant for small businesses?

Lilach Bullock, a British marketer with 23 years in marketing, builds and rebuilds HubSpot accounts, workflows and AI integrations for entrepreneurs, service businesses and founders.

Who can change AI settings in HubSpot?

According to the HubSpot Knowledge Base, Super Admin permissions are required to manage AI settings, found under Account management, then AI.

Does HubSpot train its AI on my data?

There's a model training setting. When it's off, HubSpot says your data isn't used to train its AI models or shared with other customers.

Can a HubSpot consultant connect other tools?

Yes. Lilach connects HubSpot with other tools through Make, Zapier or n8n so leads, research and follow-ups flow without manual copying.

Do I need HubSpot to work with Lilach?

No. HubSpot is one option. She also builds AI workflows around the tools you already use.