Paid Strategy
What Skills Do You Need to Run B2B Paid Ads In-House?
You need five capability blocks: building campaigns in LinkedIn Campaign Manager, Google Ads and Meta Ads Manager; wiring audiences to the CRM; producing creative at volume; measuring pipeline, not just clicks; and making weekly bid and budget calls. Platform setup is learnable in weeks; bid, audience and creative judgment takes years of reps.
Key takeaways
- In-house B2B paid media breaks into five capability blocks: platform build, audience and CRM plumbing, creative throughput, measurement, and weekly optimization.
- Platform mechanics like campaign structure, UTM building and audience uploads are learnable in weeks; bid and budget judgment calibrated to a B2B sales cycle takes 1 to 2 years of reps.
- LinkedIn Matched Audiences need a minimum of 300 matched members per campaign and the Insight Tag installed sitewide before audience and conversion data work at all.
- Creative and weekly optimization are usually better solved with software or a flat-fee agent than a new hire, since the skill ceiling is lower than the throughput problem.
- Moving ad accounts off an agency breaks tracking ownership and audience access first, so negotiate asset-copying access and map UTM conventions before relaunching campaigns.
Running B2B paid ads in-house is not one skill, it is five. Most career-site lists blur them into generic "PPC skills" and skip the part that actually determines whether a lean team can pull this off: which of those five jobs needs a trained human, which needs software, and which is cheaper to leave with an agency.
The five capability blocks behind in-house B2B paid media
Every B2B paid media program, whether it runs on one platform or three, needs the same five jobs done every week:
- Platform build: setting up campaigns, ad groups or ad sets, targeting and bids inside LinkedIn Campaign Manager, Google Ads and Meta Ads Manager.
- Audience and CRM plumbing: turning Salesforce or HubSpot records into matched audiences, custom audiences and lookalikes, and keeping those lists synced.
- Creative throughput: writing and designing enough ad variations, by audience and format, to actually test something.
- Measurement: wiring UTMs, offline conversions and CRM attribution so spend ties to pipeline, not just clicks.
- Weekly optimization: reading performance and moving bids, budgets and creative before the week is wasted.
A one- or two-person team doesn't have to be expert at all five. It has to know which ones are worth the ramp time and which to buy.
LinkedIn Campaign Manager: audience, bid and format skills
LinkedIn's main learning curve is audience construction, not campaign structure. Matched Audiences let you build targeting from a Salesforce or HubSpot list, a LinkedIn audience, or retargeting from people who engaged with a specific ad. According to LinkedIn's Matched Audiences best practices guide, the minimum campaign audience size is 300 members, and location is a required targeting facet that can shrink a segment below that floor.
Two skills matter most here:
- List hygiene for company and contact uploads. LinkedIn's help center notes that when uploading a company list you should provide the company name and the LinkedIn Company Page URL to improve match rate and accuracy. A CRM export with clean company names and URLs matches better than a messy one.
- Retargeting source selection. You can build Matched Audiences from people who visited a page, engaged with a single image, document or video ad, opened or submitted a Lead Gen Form, or engaged with your Company Page or responded to your LinkedIn Event, each pulling a different stage of buyer into the same funnel.
Two platform rules trip up almost every new in-house manager. First, website retargeting needs the LinkedIn Insight Tag installed before any data accrues, and the Insight Tag must sit in the global header or footer of the site, not just one landing page. Second, LinkedIn will only match members who have opted in across the EU, EEA and Switzerland, which can result in lower audience counts than the CRM list you uploaded, so don't plan international ABM around a 1:1 match rate.
Google Ads: match type, bidding and Performance Max skills
Google Ads is heavier on structural decisions. A new in-house marketer has to split Search campaigns by intent (branded, competitor, category), choose match types deliberately instead of defaulting to broad, and build a negative keyword list before launch rather than after wasting spend. Bidding is its own skill: moving to a Smart Bidding strategy like Target CPA or Target ROAS only works once there is enough conversion volume for the algorithm to learn from, and B2B accounts often don't hit that volume for weeks.
Performance Max adds a different difficulty. It pools Search, Display, YouTube, Gmail and Discover inventory behind one automated campaign, so the manager's job shifts from picking placements to feeding it the right conversion goals and asset groups, then watching for channels that quietly eat budget without producing pipeline. That diagnostic judgment, not the setup, is what takes time to build.
Meta Ads Manager: Advantage+, Custom Audiences and Conversions API skills
Meta's B2B use case is usually narrower than Google's or LinkedIn's: retargeting website visitors and matched CRM lists, plus lookalikes built from closed-won accounts. The core skills are building Custom Audiences from a CRM customer file, layering Advantage+ automated targeting on top instead of hand-picking placements, and setting up the Conversions API so server-side conversion events keep reporting accurate as browser tracking degrades. None of this is conceptually hard. What takes longer is judging when an Advantage+ campaign is finding real buyers versus just cheap clicks, which small B2B conversion counts can take several cycles to tell apart.
Which of these skills take weeks, and which take years
This is the split that generic PPC skill lists skip, and it's the one that should actually drive a hiring or buying decision.
| Skill | Time to competence | Why |
|---|---|---|
| Campaign and ad set structure, naming conventions, UTM building | Days to 2 weeks | Mechanical, documented, same steps every time |
| Matched Audience, Custom Audience and list uploads | 1 to 3 weeks | Format and hygiene rules are learnable from platform help docs |
| Writing ad copy and basic creative variations | 2 to 4 weeks for competent, months for consistently good | Requires knowing the offer and ICP, not just the ad unit |
| Bid and budget judgment calibrated to a real B2B sales cycle | 1 to 2 years of reps | Needs enough closed-won and closed-lost cycles to know what "working" looks like |
| Diagnosing why a campaign underperforms (audience, offer, creative or tracking) | 1 to 2 years | Pattern recognition across many campaigns, platforms and failure modes |
The pattern: anything that's a documented platform setting is a weeks-long skill. Anything that requires judging a tradeoff against a sales cycle you can't see in a dashboard takes years. That's the real argument for keeping an experienced hand, whether in-house, on an agency retainer, or built into software, on the judgment calls while a newer hire or a tool handles the mechanics.
What CRM and marketing ops skills does paid media depend on
Paid media in B2B is only as good as the CRM data feeding it. Three specific skills matter:
- UTM discipline. Every campaign, ad group and even creative variation needs a consistent UTM taxonomy that maps cleanly to a CRM field (usually a custom field on the Lead or Contact record, or a Campaign Member status in Salesforce), so a form fill can be traced back to the exact ad that drove it.
- Offline conversion and list sync. Closed-won deals, sales-qualified leads and pipeline stage changes need to flow back into the ad platform, either as an offline conversion import or a refreshed audience list, so bidding algorithms optimize toward revenue signals instead of just form fills.
- Segment and report building. Pulling a closed-won list, a target-account list, or an active-pipeline list out of Salesforce or HubSpot in a clean, exportable format is a marketing ops skill, not a paid media skill, and it's usually the actual bottleneck on audience quality.
Most in-house teams underestimate this block. A marketer who's excellent at bidding but can't pull a clean closed-won list out of Salesforce will always target worse audiences than a mediocre bidder with great CRM access. See our guide on turning CRM data into a paid media plan for the mechanics.
How much creative and copy throughput does one marketer actually need
Running three platforms well means running several audiences on each, and each audience needs more than one message and format to learn anything from testing. Two creatives per campaign is not a test, it's a coin flip. A single marketer covering LinkedIn, Google and Meta for even two or three segments needs a steady pipeline of image, carousel and short video variations, refreshed often enough that an audience doesn't go stale, which on LinkedIn's smaller B2B audiences happens faster than on consumer platforms.
This is the block where a solo or two-person team hits a wall first, not because the skill is hard to learn, but because design and copy output takes hours per variation regardless of skill. Most lean teams solve this with templated, on-brand creative systems rather than a dedicated designer. We cover the specific approach in how to produce ad creative without a designer.
Hire, buy software, or keep an agency: a decision table by capability block
Not every gap should be filled the same way. Here's how to think about each block honestly.
| Capability block | Hire when | Buy software when | Keep an agency when |
|---|---|---|---|
| Platform build | Spend is large enough across 3+ platforms to need daily hands-on management | You need staged, reviewable campaign builds without a full-time operator | You lack any in-house paid media literacy yet |
| Audience and CRM plumbing | You have a dedicated marketing ops hire who owns the CRM already | You need repeatable CRM-to-audience sync without new engineering work | Rarely the agency's strength unless they have real CRM access |
| Creative throughput | Volume is high enough to justify an in-house designer's full time | You need many on-brand variations fast without a design hire | Agency creative is usually slower and less on-brand than a templated system |
| Measurement | Attribution complexity spans many tools and needs a dedicated analyst | You need UTM and offline conversion workflows without manual spreadsheet work | Agencies rarely own your CRM data model |
| Weekly optimization | Spend and campaign count justify daily monitoring | You need consistent weekly bid and budget recommendations without the headcount | Good for teams that want full hands-off management, at the cost of control |
Tofu Ads Agent is one option in the "buy software" column across audience building, creative and weekly optimization: it reads the CRM and ad accounts to build the ICP and plans, builds audiences from a Salesforce or HubSpot list or a plain-English description, produces creative in the brand's own fonts and templates, and stages campaigns into the team's own ad account switched off so a person still approves every launch. It's a flat $2,500 a month, not tied to ad spend, with a recommended minimum media budget of $3,000 to $5,000 a month, so it fits teams that have some budget but not the headcount. For a fuller breakdown of hire-vs-buy-vs-agency tradeoffs, see our PPC agency alternatives decision guide.
What breaks when you bring ad accounts in-house from an agency
Pulling ad accounts back from an agency fails more often on plumbing than on strategy. The usual casualties:
- Tracking ownership. If the agency's team installed the LinkedIn Insight Tag, Google tag or Meta pixel through their own tag manager container, conversion tracking can break the moment access changes, since the Insight Tag has to be present in the site's global header or footer for Matched Audiences and conversion data to keep accruing.
- Audience access, not just portability. LinkedIn's own best practices note you can use asset copying to move website retargeting, company list or contact list audiences across Campaign Manager accounts, but that only works while someone still has access to both accounts to copy from. Negotiate that access before the agency revokes it, not after. Engagement-based retargeting audiences, built from ad views or Lead Gen Form activity, aren't part of that copyable set, so plan to rebuild those from CRM or ad data, and remember LinkedIn requires a minimum campaign audience size of 300 member accounts, so a freshly rebuilt list needs to clear that floor before it delivers.
- Negative keyword and exclusion lists. These live at the account or campaign level and rarely get documented well enough to hand over cleanly.
- UTM and naming conventions. If the agency used its own taxonomy, historical CRM reporting on campaign source breaks the day you switch to a new one, unless you map the old convention first.
- Learning phase resets. Smart Bidding and Advantage+ style automated bidding need a fresh run-in period in a new account structure, so expect a dip before performance stabilizes.
The fix is sequencing: confirm admin access and audience-copying or tag ownership first, rebuild what can't be copied from CRM source data before touching campaign structure, map the UTM taxonomy before relaunch, and expect a short dip while automated bidding relearns. Teams that skip this and just "flip the switch" are the ones who lose a quarter of performance data in the handoff.
Frequently asked questions
Can one marketer realistically run LinkedIn, Google and Meta ads alone?
Yes, for a lean budget and a small number of segments, but it works best when that marketer buys the gaps in creative and weekly optimization rather than trying to do all five capability blocks solo. The platform build and CRM plumbing skills are learnable in weeks; the time cost is in creative volume and in the judgment that takes years of reps.
How long does it take to get good at Google Ads bidding for B2B?
Basic campaign structure and match type decisions take a couple of weeks to learn. Calibrating bid strategy, budget pacing and Smart Bidding goals to an actual B2B sales cycle, where conversions are sparse and delayed, typically takes 1 to 2 years of running campaigns through full sales cycles.
Do I need a designer to run B2B paid ads in-house?
Not necessarily a dedicated hire, but you do need a repeatable way to produce several on-brand creative variations per audience every week, since two creatives isn't enough to test anything. Many lean teams solve this with templated creative systems instead of a full-time designer; see our guide on producing ad creative without a designer.
What's the minimum team size to bring B2B paid media fully in-house?
There's no fixed headcount, but a one- or two-person team can cover platform build and CRM plumbing themselves if they offload creative throughput and weekly optimization to software or a part-time agency relationship, since those two blocks scale with volume, not with strategic skill.
How do I keep LinkedIn conversion tracking and audiences intact when switching from an agency?
Confirm who owns the LinkedIn Insight Tag before you change access, since it has to sit in the site's global header or footer for Matched Audiences and conversion data to keep accruing. Negotiate asset-copying access for website, company list and contact list audiences before the agency revokes access, and plan to rebuild engagement-based retargeting audiences from CRM or ad data since those aren't part of the copyable set.
Does marketing ops or demand gen own the paid media skill set?
Both, but for different blocks. Demand gen typically owns platform build, creative and weekly optimization, while marketing ops owns the CRM plumbing, UTM taxonomy and offline conversion sync that paid media depends on. Teams that treat these as one role often end up weak on the CRM side.
When does it make more sense to keep an agency than hire in-house?
Keep an agency when you have no in-house paid media literacy yet and need someone to own strategy and platform relationships while you build internal capability. It's a weaker fit once you need tight CRM integration, since most agencies don't have deep access to your Salesforce or HubSpot data model.
Sources
- Matched Audiences best practices · LinkedIn Help
- LinkedIn Matched Audiences · LinkedIn Help
- Upload company and contact targeting lists for LinkedIn Ads · LinkedIn Help
- Create a Matched Audience from your Audiences page in Campaign Manager · LinkedIn Help
