Construction Project Leads Databases Compared for Outbound

A construction project leads database is a source of project records. Each record shows what is being built, where, at which stage and for roughly what value. It also names the firms attached to it: the developer or owner, the architect, the engineer and the contractor.

Four source types matter for outbound: North American commercial databases (Dodge and ConstructConnect), UK project databases (Glenigan and Barbour ABI), public planning portals, and building permit data.

Choose by timing. If your product has to be specified, buy the source that shows the planning and design stage. If you sell to whoever wins the job, buy bid and award data. Use free public data as a support layer.

The database matters less than the list you build from it. Take one client of ours as the example: a UK manufacturer of fall-protection equipment for construction sites. Our cold email and LinkedIn work for them, over 2 years, brought 600+ leads and 200+ sales calls across 24 countries, and AECOM, bp, Skanska, Doka, Altrad and Saipem replied. More campaign results are in our case studies.

When I review a campaign, I open the spreadsheet of names before I read a line of copy, because copy cannot fix a bad list. I explain how I take a list apart in my teardown of a bad B2B lead list.

What is a construction project leads database?

It is a tracker for jobs. It follows a building or a civil works contract from an early idea to a signed contract and records which firms are involved at each step.

Each type does a different job in construction outbound.

Of everything a record can hold, outbound needs three things:

1. Timing. A stage field you can filter on, so your first message lands while the decision is still open.
2. Firms by role. The developer, the designers and the contractor, each named, because each one buys at a different moment.
3. A path to a person. A named decision maker, or enough detail to find one.

Timing alone is a research tool. Timing plus firms plus a person is a list you can send to.

Before you open any database, go back to your own customers, past and current, and work out who signed off on each project you won. Write down their exact job titles and the stage each one decided at. Those are the titles and the stage you filter for, not a generic "manager".

I start every list this way, with a client's past and current customers first and similar firms after that. Skip it, and you will scroll through endless records waiting for one to look familiar.

Which project data sources matter most for outbound?

There are two pairs, split by geography. Dodge and ConstructConnect cover North American commercial work. Glenigan and Barbour ABI cover the UK, and Glenigan also covers the Republic of Ireland. Everything in this table comes from each vendor's own official page so read the counts as vendor claims.

Read the table in two passes. Start with geography: if your buyers are in Texas, the UK pair is irrelevant, and the same goes the other way. Then look at stage: which vendor shows projects at the moment your buyer decides.

A big count tells you what the vendor tracks. It says nothing about your buyers. The number that matters to you is a slice of it: your trade, your region, your stage, this quarter. Ask for that slice in writing before anyone shows you a demo dashboard.

What does each database usually contain?

Every vendor sells "project data", but for outbound only a short list of fields is useful. This is what each official page mentions.

Two caveats apply to the contact column.

Contacts follow the stage. At planning, a record may show an applicant and an architect and no contractor, because nobody has been appointed yet. A blank contractor field at that stage is normal. It tells you who to message now and who to wait for.

A named contact is only a starting point. People move between firms and between projects, and a name tied to a job last year may now work somewhere else. The checks you would run when auditing any lead list before sending apply to project exports too: titles, seniority, freshness and verification.

Dodge vs ConstructConnect for US project leads

Both cover US commercial work, and both say they follow projects from early design onward. The difference is what each one leads with.

Dodge's pitch is time. It talks about seeing projects before they reach the bid boards, with planning-stage records, specs and contacts. ConstructConnect's pitch is the bidding workflow: following a job to contract award and seeing which firms bid and which one won.

That gives a simple choice rule:

• Your product gets written into drawings or specifications (curtain wall, roofing systems, mechanical equipment): you need the early signal. Start with planning-stage records and contact the developer, the architect and the engineer while the specification is still open.
• You sell to whoever wins the job (subcontract trades, materials delivered to site, plant and equipment hire): you need the bid-phase workflow. Bidder and award data tells you which contractor to contact and when.
• You sell both ways: send the same sample request to both vendors for one metro area you know well, and compare which export gives you more named people at the stage you need.

The expensive mistake is a stage mismatch. I see it in almost every construction list I review. If you message a general contractor about a job still at planning, nobody there owns it yet. If you message the architect after award, the specification is usually fixed.

Glenigan vs Barbour ABI for UK construction leads

Both are UK project databases, and their official pages point to different strengths.

Glenigan leads on earliness. It says an exclusive partnership with the Planning Portal gives it projects before local authorities see them, and that it also tracks projects outside the planning process. It covers the Republic of Ireland as well as the UK, and it lists LinkedIn profiles among its decision-maker contact details. If your outbound opens on LinkedIn, that field saves you a research step.

Barbour ABI leads on bidding history. It describes verified bidding history and direct contact details, exports leads to a CRM, and sends daily or weekly alerts on saved searches.

The UK choice rule:

• You need to be first at planning, or you also sell into Ireland: Glenigan.
• You need to know which contractors keep winning the type of job you supply: Barbour ABI.

A contractor that keeps winning warehouse or school jobs in your region is a target in its own right, whatever project happens to be live this month. That turns a project feed into an account list you can work for a year.

Can public planning and permit data replace a paid database?

For most suppliers, no. Free sources are good at early signal and geography and weak at the thing outbound needs most, which is a named person. Treat them as support data.

• England's planning data. Planning Data is an official platform for planning and housing datasets in England. It shows datasets on an interactive map, offers bulk downloads and has an API. Expect planning records rather than ready decision-maker contacts.
• US permit data. The Building Permits Survey collects data every month from roughly 8,400 places that issue permits and covers privately owned residential structures. Its main data items are units authorized, number of buildings and total valuation. That shows you where residential building is growing. It is aggregate data by place, so it gives you no project owner to contact.

Use this decision rule:

How do suppliers turn project records into an outreach list?

Before you set any filter, ask whether your market is big enough for project signals to matter. When the addressable market is large, timing signals are worth the work. When it is small, I skip them. Picture 2,000 target companies with live signals on about 10 of them: chasing those 10 leaves everyone else untouched, so the job becomes an offer that works for the whole market.

In construction this split is real. A manufacturer selling to a few hundred specialist contractors in one country can usually reach all of them, so project data works as a priority flag on an account list. A materials supplier selling across a continent cannot reach everyone, so project records become the filter for which firms get a message in the next few weeks.

Even then, filtering only fixes quality, and you need quantity as well. With 50 to 100 perfect-fit records a month you are flipping a coin, because plenty of those people will be busy, not ready yet or simply not interested. Enough volume evens out that luck.

Then work through it in order:

1. Filter the records. Keep the geography you can serve, the stage where your buyer decides, a value band that fits your deal size, the project types your product belongs in, and start dates inside your lead time.
2. Split each record by role. Developer, designers, main contractor, and the subcontractors once they are known.
3. Map real titles for each role. Use the titles that actually sign for your product in that niche, and exclude generic ones such as Assistant, Associate and a bare "Manager". A qualified title like Commercial Manager stays in.
4. Turn one record into a contact cluster. Pick several people across the firms on the job, so one person on holiday does not end the opportunity.
5. Add lookalikes from your own history. Contractors and developers who bought from you before are likely to be on similar jobs now.
6. Build the people layer properly. Use provider tools such as Sales Navigator, Apollo or ZoomInfo for names and titles.
7. Prepare the sequence. See the next subsection.

A worked example: one UK warehouse record

This record is invented for illustration. Say you manufacture loading dock equipment and sell in the UK. An alert brings in a record like this:

• New-build distribution warehouse, East Midlands
• Stage: planning application submitted
• Value: the vendor's estimate, which you check against your deal size
• Firms: a logistics property developer (the applicant), an architecture practice, a structural engineer
• Main contractor: not yet appointed

That record becomes a target list in two waves.

Leave out assistants, coordinators, marketing staff and anyone whose title says nothing about the project. Add the structural engineer only if your product affects the structure.

The result is one record and a handful of named people across 3 firms, contacted at two moments. Then check which contractors have won similar warehouses in the region, using bidding history where you have it, and add them as accounts before the award is announced. For the people layer in full, see how to build a B2B lead list worth sending to.

Getting the list ready to send

In the first message, sell a short conversation about this job, not your product catalog. Name a similar project you supplied and what it delivered, because past clients and results belong inside the message.

Plan at least 5 to 6 touches, ideally across email and LinkedIn, since one or two messages are never enough. Expect prospects to check you before they answer: your LinkedIn profile, your website, your project references, and a thin profile or empty site costs you the reply.

If you are unsure which channel should open, read whether to start with cold email or LinkedIn before you load the first batch.

What should you check before you buy a database?

Run these checks before you sign or export anything:

1. Sample rows from your own slice. Your territory, your stage and your project types. Ignore the showcase rows the vendor picks for you.
2. Fields on the sample. Stage, value, location, firms by role, named contacts, and a last-updated date on each record.
3. Definitions. What the vendor counts as an "active" project and as a "contact record", and how often records update.
4. Contact reality. Pick 20 sample records, find the decision maker you would actually write to, and count how many you can name with confidence.
5. Verification. Check the sample emails with NeverBounce, DeBounce or a similar tool, and treat a high invalid share as a warning about the whole export.
6. Delivery. Check that records can reach your CRM or campaign tool without anyone retyping them.
7. Cheap add-ons. If someone offers to "add emails" to your export for a low flat fee, expect low open rates and high bounces.
8. Price for your slice. Get a written quote per seat and per region, and set it against what one good project is worth to you.

Decide before you look at the sample what counts as a pass. If most of your 20 records end with a named, verified decision maker in the role you sell to, the database can feed outbound. If few do, keep it as a research tool and build the people layer somewhere else.

Stage first, then named people

Pick the source by stage. Then judge it by how many named people you can reach in your own slice.

Choosing between project databases and want a view on turning one into a sendable list? Book a call and I will take a look with you.