Agency or in-house marketing? What shapes your growth curve
Agency vs in-house marketing, decided on the growth curve instead of salary vs fee: time to reliable data, CAC, brand compounding, retention and ownership risk.
TL;DR
Agency or in-house is not settled by putting a salary next to a fee. The growth curve is the product of five factors: time to reliable data, the slope set by customer acquisition cost and conversion, compounding from brand and organic channels, the second half carried by retention, and the drawdown when a key person or ownership is lost. Each model wins different factors. Accounts and data should stay with the company in every model.
The agency vs in-house marketing decision is the choice between running marketing with your own employees, with an outside specialist team, or with a division of work between the two. It is usually made by comparing a salary with a proposal, yet its effect shows up in revenue over time.
This post takes that curve apart factor by factor. The rest of the series goes deeper on cost, measurement, brand and the hybrid model.
What are the three ways to run marketing?
Fully in-house. Strategy, ads, content, design, the website and email sit with employees. Product knowledge and decision speed stay inside, and so does the work of hiring and managing every discipline.
Specialist agency. An outside team holds six disciplines together, without six hires. Transferring product knowledge and protecting ownership by contract become the company's job.
Hybrid. Strategy, brand ownership and the customer relationship stay inside; specialist execution and automation come from outside.
In 2023, 82% of ANA members had an in-house agency, up from 42% in 2008, yet 92% of respondents also worked with external agencies, and those companies did an average of 61% of their work in-house (ANA, 2023). The sample is US-centric and mostly large companies, but the lesson travels: the question is which work goes where.
In 2025, Türkiye's e-commerce volume grew 52.2% in nominal terms and exceeded 4.57 trillion TL, 19.3% of general trade (Turkish Ministry of Trade, 2026). Nominal TL growth is not real growth, but at that size, how marketing is organised is no longer a detail.
What shapes the growth curve?
Five factors draw the revenue line over time:
- Starting point: when decisions begin to rest on reliable data.
- Slope: new customers per unit of spend, set by customer acquisition cost (CAC) and conversion rate.
- Compounding: what brand and organic channels keep delivering after spend stops.
- Second half: retention, the repeat purchases after the first order.
- Drawdown risk: how far the curve falls when a key person, an account or the data disappears.
The factors multiply. Excellent creative on broken measurement trains the ad algorithm on the wrong signal, and the slope flattens. A low CAC with no repeat buyers means every month starts from zero. When one factor nears zero, strength elsewhere cannot make up for it. Choosing a model means deciding who carries each factor best.
How fast does each model reach reliable data?
The curve lifts when the data becomes trustworthy, and that is often a skills problem. In Gartner's 2024 survey of 378 senior marketing leaders, only 52% said they can prove marketing's value and receive credit for it; three of the biggest barriers were talent-related, including analytical talent and the talent to integrate and analyse data (Gartner, 2024).
An in-house build often spends most of its first 90 days hiring while measurement waits behind ads and content; a specialist team that has built the setup many times can have it running within weeks.
The ad platforms' learning period applies under every model: bidding needs a volume of conversions before delivery settles, and we plan for 30 to 45 days, as described in performance ads for Turkish brands entering Europe. What an agency can do is avoid stretching it with broken tracking or early budget jumps. The measurement layer is in measurement, conversion and retention.
How do CAC and conversion set the slope?
CAC is the total spent to win new customers in a period, people and tools included, divided by the new customers won. It moves from two sides: the numerator shrinks or the denominator grows.
For example, a brand spends $10,000 a month, the site gets 20,000 visits and 1.5% of visits become orders: 300 orders, and if all are new customers, a CAC of about $33. Simplify checkout, speed up the pages and lift conversion to 2% on the same traffic, and orders rise to 400 while CAC falls to $25. Same budget, a quarter lower CAC.
Higher conversion pays twice: more conversion signals help the ad platform find likely buyers, so it can acquire customers more cheaply on the same budget. The reverse holds too: one creative run for weeks pushes frequency up, click-through down and cost per result up. Holding the slope takes regular creative refreshes and a testing cadence.
Here the difference is bandwidth. In many in-house teams the person running ads cannot touch page speed or checkout; that waits in a developer's queue. Our design, engineering and ad management share one table with no account-manager layer, so a conversion change is settled in one meeting. Cost lines in full: the real cost of an in-house marketing team.
How do brand equity and organic channels compound?
Paid ads work like rent: stop paying and the traffic stops. Brand and organic channels work more like interest. A guide that ranks keeps bringing visits in months with no spend, and people who already know the brand click and buy with less hesitation, which pulls paid CAC down. Each month's contribution sits on top of the last.
The trap is timing: the effect appears months later, so a team under quarterly pressure moves budget to whatever can be measured this week. In Gartner's 2025 CMO spend survey, budgets stayed flat at 7.7% of company revenue and 59% of leaders called theirs insufficient for their strategy (Gartner, 2025). Its 402 CMOs and marketing leaders in North America, the UK and Europe mostly work at companies above 1 billion USD in annual revenue. These are large-company benchmarks; 7.7% is not a target for a smaller business.
Content and SEO depend on continuity and are the first work a busy team postpones. With an agency, the voice can drift if it is shaped entirely outside. Keep the owner of the voice inside and the cadence in a system; more in brand equity and organic growth. How that organic work gets cited in AI answers is covered in generative engine optimization.
What do retention and automation add to the curve?
New customers draw the first half of the curve; returning customers draw the second. If a customer buys 1.6 times in the first year instead of 1.2, and profit per order holds, each acquired customer brings a third more gross profit: faster growth at the same CAC, or room for a higher one.
Most retention work repeats: order confirmation, shipping updates, a review request, a win-back message. When these flows are missing, it often goes unnoticed, because missing revenue never shows in a report. ANA respondents name managing workflow as projects multiply, scaling efficiently and prioritisation as their biggest in-house challenges (ANA, 2023); in our experience, the same pressure pushes retention flows down the queue.
We hand the repeating part to four robots: content, ads, communication and operations. The communication robot sends lifecycle messages by consent status. See a marketing automation stack for a five-person team.
What are the key-person, knowledge and ownership risks?
A growth curve can also fall, and the usual drops come from organisation, not from ads:
- Key person. One employee runs the ad accounts and leaves; the replacement does not know the history, the rules, or why a test was stopped.
- Knowledge. Campaign rules, the brand voice and the tracking setup live in someone's head.
- Ownership. The ad account or tracking setup sits under the agency's business account. When the agency changes, trained campaigns and history leave with it, and learning starts again.
All three happen under every model and are managed the same way: written process and accounts in the company's name.
Outsourcing work is not outsourcing ownership
Ad accounts, the tracking setup, the domain, social profiles and the customer database should be opened in your company's name, with the agency given access only. Write into the contract that code, content, accounts and documentation transfer to you on exit. In our contracts, all four stay with the client.
Agency vs in-house marketing: nine factors, three models
| Factor | Fully in-house | Specialist agency | Hybrid |
|---|---|---|---|
| Time to reliable data | Depends on hiring, often slow | Fast; learning period still applies | Built outside, owned inside |
| Breadth of disciplines | One hire per discipline | Several disciplines in one team | Critical ones inside |
| Product and customer knowledge | Deepest | Needs transfer | Stays inside, goes out as briefs |
| Conversion testing cadence | Limited by developer time | High with design, engineering and ads together | Decided inside, built outside |
| Brand and organic compounding | Postponed under pressure | Steady, voice can drift | Voice inside, cadence in a system |
| Retention and automation | Often queued | Setup and upkeep, if in scope | Rules inside, automation outside |
| Cost structure | Fixed salaries, even at low volume | Contracted monthly fee, ad budget separate | Small fixed core plus a fee |
| Key-person risk | High | Managed by contract | Low; knowledge inside and written |
| Account and data ownership | Company by default | Depends on the contract | Company; agency part by contract |
The rows in-house wins are real: product knowledge and natural ownership are hard to match, and when volume fills a full-time role in every discipline, in-house is often cheaper too. The hybrid marketing model post covers where each piece of work belongs.
How do you decide between an agency and an in-house team?
The decision between an agency and an in-house marketing team comes down to three tests: volume, measurement and leadership. In-house is the stronger choice when the work fills a full-time role in each of six disciplines (brand and creative, web and e-commerce, performance ads, social media, SEO and GEO, marketing automation), when sales can be read from the company's own order data, and when a senior marketing lead can run the team. When volume is lower, measurement is not in place or no senior lead exists, a specialist agency usually lifts the growth curve sooner, because one team covers several disciplines without six hires. Between the two, a hybrid model keeps strategy, brand ownership and the customer relationship inside and buys specialist execution outside. In every model, the ad accounts, the tracking setup, the domain and the customer database should be registered in the company's name, with the agency given access only.
Six questions put those three tests to work:
- Does the work fill a full-time role in each of six disciplines: brand and creative, web and e-commerce, performance ads, social media, SEO and GEO, marketing automation?
- Can you read sales from your own order data rather than an ad dashboard?
- Do you have a senior marketing lead who can set strategy and run the team?
- Are lifecycle emails set up and measured?
- Are the ad accounts, the tracking setup and the domain in the company's name?
- If one person left tomorrow, what knowledge would leave with them?
Yes to the first three makes in-house a strong candidate. No to the second and third means bringing in specialists usually lifts the curve sooner. In between, hybrid is often the lowest-risk start. Each no to the last three is a gap to close first, whichever model you choose.
What does it cost to start with Guardino Technologies?
For a company building its marketing from scratch, the start is the Launch Program: eight weeks from contract to go-live for $9,900. Payment is 40% at contract, 30% at design approval and 30% at go-live. It covers the web and e-commerce platform, measurement, the first automation flows and social media setup. The monthly programs share one scope (social media engine, Meta ad management, maintenance): Monthly A is a flat $1,290 a month, Monthly B is $990 a month plus 4% of net online revenue. Monthly programs run on a 12-month term. You pay ad budgets directly to the platforms, with no margin hidden in media. Add-on pricing is on the programs page.
The rhythm is fixed: a report every Monday at 08:00, a monthly strategy meeting, a quarterly review. See the process and about pages, or start from performance ads, web and e-commerce, marketing automation or the full services list.
The brief takes about 10 minutes; a written summary follows by email, and the contract and payment link within 1 business day. If in-house fits you better, we will say so.
Also available in Türkçe.
