Skip to main content

Should you spend on your own social ads or on influencers? What an Australasian Marketing Journal study found

Most brands now pay for two kinds of social post: ads in their own name and posts from creators. A survey of online shoppers and a real TikTok launch of two Finnish energy drinks suggest they are not rivals. Each is better at a different job.

Kumar et al. find a brand’s own paid social posts win more engagement, while influencer posts sell more, in a 1,050-person survey and a TikTok drinks launch.
Date

7 November 2026

Marketing in 3 · 2024 in research · No. 17 · Paper published 23 April 2024

A young woman holds up a phone on a gimbal to film on a busy street at night, under glowing red and orange neon signs
A creator reaches fewer people than a paid ad, yet those views kept closer pace with sales. Photo: Sergei Bezzubov on Unsplash.

The short answer

It depends on the job. A survey of 1,050 online shoppers and a two-month TikTok launch for two Finnish energy drinks gave the same split. A brand's own paid posts drew more engagement, such as clicks and interest, while posts from influencers did more to move buying. Run in the same weeks, each still did its own job, and both lifted sales in shops the campaign never linked to.

Evidence level
  1. One study
  2. Several studies
  3. Meta-analysis

Key takeaways

  • In the survey, a paid brand ad lifted engagement about 62% more than identical content said to come from an influencer people followed.
  • For purchase intent the order flipped: the creator version raised it about 75% more than the paid brand ad.
  • On TikTok, each 10% rise in daily influencer views went with roughly 1% more units sold; the same rise in brand-ad views went with about 0.3%.
  • Paid brand ads were seen far more often, about 32,400 times a day against about 3,100 for the creators' posts, and were clicked at 0.20% against 0.15%.

What the researchers did

Ashish Kumar, Daniel Rayne and Ching Sophia Yiu of RMIT University in Melbourne, with Jari Salo of the University of Helsinki, took on a question behind almost every social media budget. A brand can pay to put its own posts in people's feeds, or it can pay creators to post about it. Most brands do both, yet earlier research had mostly judged each route on its own. The team set the two side by side on two outcomes: engagement and buying.

The first study was an online survey of people who shop online, use social media and follow at least one influencer. They were split into three equal groups. One group was only told that a brand was launching a new product. A second saw the launch as a paid ad in their feed. The third saw the same content but was told it had been shared by an influencer they follow.

The second study moved to a real launch. Soul Maté, a Finnish food and drinks company, introduced two energy drinks on TikTok. For two months it ran paid, targeted ads under its own name while two hired creators posted from their personal accounts. Every post pointed to the same online shop.

Study 1
Online survey of 1,050 shoppers, 350 per group; average age about 43, close to half of them women
Groups
No ad; a paid brand ad in the feed; or the ad's content presented as a post from a creator they follow
Study 2
Two-month TikTok launch of two energy drinks, Passion Guava and Ginger Lemon, by the Finnish firm Soul Maté
Creators
Two hired influencers with more than 82,000 followers, chosen and briefed using four focus groups
Measured
Stated engagement and purchase intent in Study 1; daily views, clicks and units sold through one online retailer in Study 2
Analysis
Bayesian models; both drinks pooled, with sales modelled from each day's views of each kind of post

Which kind of post gets attention, and which gets the sale?

Compared with people who saw no ad at all, both kinds of post raised engagement and buying intentions. They did not raise them equally. The paid brand ad produced the top engagement score of the three groups, while the influencer version produced the top buying score.

Neither a person's age nor their schooling shifted these results. Gender did: the creator version had a stronger pull on women, for engagement and for buying alike.

The authors explain the split with two ideas. Followers build a one-sided friendship with creators they watch, and see them as people like themselves, so a creator's tip feels like advice from someone they know. A brand's own post is easier to read as a sales pitch. Yet it keeps the brand talking to people directly, with no third party in between, and that seems to be what builds engagement.

Average survey scores by what people saw
Average survey scores by what people saw01234Engagement, no ad: 1.91 Engagement, no ad1.91Engagement, brand ad: 3.52 Engagement, brand ad3.52Engagement, influencer post: 2.76 Engagement, influencer post2.76Purchase intent, no ad: 1.63 Purchase intent, no ad1.63Purchase intent, brand ad: 2.91 Purchase intent, brand ad2.91Purchase intent, influencer post: 3.12 Purchase intent, influencer post3.12
Average survey scores by what people saw
Group Value ()
Engagement, no ad 1.91
Engagement, brand ad 3.52
Engagement, influencer post 2.76
Purchase intent, no ad 1.63
Purchase intent, brand ad 2.91
Purchase intent, influencer post 3.12

Study 1, online survey, 1,050 shoppers in three groups of 350; averages from Table 1; scale range not given in the main text.

Did the split hold when real money changed hands?

It did. On TikTok the paid ads reached many more people, because the brand could bid for placement, while the creators' posts reached mainly those who visited their feeds. Across the launch, the online shop sold an average of 257 units a day.

Because every post used one shared link, nobody could see which post produced which sale. The team instead modelled each day's clicks and sales against that day's views from each source. Views of brand ads were tied more closely to clicks, about 2.5 times as strongly as views of creator posts. For sales, the order reversed, and the gap was wider still.

TikTok launch: how strongly clicks and units sold tracked each kind of view (elasticity)
TikTok launch: how strongly clicks and units sold tracked each kind of view (elasticity)00.020.040.060.080.10.12Clicks, brand-ad views: 0.100 Clicks, brand-ad views0.100Clicks, creator-post views: 0.040 Clicks, creator-post views0.040Units sold, brand-ad views: 0.033 Units sold, brand-ad views0.033Units sold, creator-post views: 0.098 Units sold, creator-post views0.098
TikTok launch: how strongly clicks and units sold tracked each kind of view (elasticity)
Group Value ()
Clicks, brand-ad views 0.1
Clicks, creator-post views 0.04
Units sold, brand-ad views 0.033
Units sold, creator-post views 0.098

Study 2, two months of daily data for two energy drinks, Table 4. Read 0.10 as: one tenth more views, about 1% more of the outcome.

How to think about splitting the budget

The practical advice in the paper mirrors that split. If the goal is engagement and a lasting relationship, put money into ads under the brand's own name. If the goal is sales in the near term, creators earn their fee.

They go a step further and suggest that social media can now carry some of the awareness work once left to mass media, given that its effects reached shops the campaign never mentioned. Keep the scale in mind, though. This was one launch of two low-cost drinks on one platform, so treat the split as a strong starting hypothesis for your own tests rather than a fixed rule.

Marketing in 3 steps

Ready. Set. Market.

ReadyDecide and prepare
  • Give each post one job. Before you brief anyone, write one line per campaign: is it there to start conversations or to sell? Give conversation goals to your own paid posts and sales goals to creators.
  • Choose creators who speak to your buyers. Shortlist three to five creators whose followers look like your customers. Ask each for their audience's age and gender mix, and favour those closest to the people who already buy from you.
SetBuild it
  • Tag every post separately. Create a separate link or discount code for each creator and for each paid ad. Sales that all land in one bucket cannot tell you which post earned them.
  • Ask for a tap in every post. Close each post with a single clear action, such as a shop link, a poll or a code to copy. A tap tells you far more about buying than a view does.
MarketLaunch and measure
  • Run both routes in the same weeks. Launch your paid posts and the creator posts together over four to eight weeks, rather than one after the other, so each can support the other.
  • Check sales beyond the link each week. Every Monday, set clicks and coded sales next to sales from stores and other retailers. Move money towards the route that sells more per thousand views.

Get it by email

New issues every Monday and Thursday, in your inbox. Free, no ads, unsubscribe anytime.

Your address is used only to send Marketing in 3. You will get one email to confirm.

What to expect

  • Small reach from creators. Creator posts will likely be seen far less often than ads you bid for, so judge them on sales per view, not on total views.
  • Modest moves. The links between views and results were real but small, so big swings in views bring gradual changes in sales, not jumps.
  • Brand posts still count. Some earlier studies found a brand's own social posts losing their pull. Here they lifted both engagement and sales.
  • Discounts still work. Days with a price promotion sold more on top of the effect of either kind of post.

Where it does not work

  • One launch, one platform. Everything in Study 2 rests on a single Finnish firm and a pair of hired creators, all on TikTok, over two months.
  • Cheap, everyday products. Energy drinks are a low-risk buy. Cars, holidays or other considered purchases may favour a different mix.
  • A label, not a real creator. In the survey the influencer group saw the same content as the brand group, only described as shared by someone they follow. Buying was measured as intention.
  • Sales modelled, not traced. Both routes sent shoppers to one link, so the split of sales comes from statistical models of daily views, not from tracking each buyer.
  • Survey details sit elsewhere. The main text gives neither the country the survey ran in nor the range of its rating scales. Both sit in an online appendix that this summary could not draw on.

Questions marketers ask

Are influencers better than brand ads on social media?

Not across the board. Paid posts in the brand's name drew stronger engagement, whereas creator posts did more for buying plans and for real sales, in both a survey and a TikTok launch (Kumar et al., 2025).

Should influencer campaigns be judged on views or on clicks?

Clicks. In the TikTok data, days with more clicks predicted sales better than days with more views, for brand ads and creator posts alike, and more so for creators.

Do influencer posts only sell through the link in the post?

No. Purchases in shops and at other sellers, which the posts never linked to, also climbed with views of both kinds of post, and creators pulled harder, so link-based tracking understates their value.

Do women respond differently to influencer marketing?

In this survey, yes. Women scored higher on engagement and on plans to buy whatever they saw, and the gap in favour of creator content was wider for them. Older and younger, more and less educated respondents behaved alike.

Does running brand ads and influencers together waste money?

Probably not. Both ran in the same weeks and each kept its own positive link with clicks and sales, which the authors read as the two working together. Their models did not test for an extra boost from combining them.

Also new at the time

Journal of Marketing · 2024

No Comments (from You): Understanding the Interpersonal and Professional Consequences of Disabling Social Media Comments

Daniels & Wu

Many influencers switch off comments to shield themselves from abuse. Across seven studies, built on a Twitter dataset and six experiments, audiences thought less of creators who did so and were harder to persuade, because a closed comment box looked like an unwillingness to hear followers, which made the creator seem less sincere. The penalty shrank when people saw a good reason for self-protection. A short note on why comments are closed may soften that cost.

Journal of Interactive Marketing · 2025

Getting the Recipe Right: How Content Combinations Drive Social Media Engagement Behaviors

Lapresta-Romero et al.

Rather than testing one feature at a time, this analysis of 516 Instagram stories looked at how elements work in combination, using a method that hunts for winning mixes rather than single effects. Four different mixes reached high engagement, which the authors label by their emphasis: attention-grabbing, informative, affective and relational. There is more than one way to build a story that works, so before blaming one element for a weak story, look at the whole blend around it.

Source

Kumar, A., Rayne, D., Salo, J., & Yiu, C. S. (2025). Battle of Influence: Analysing the Impact of Brand-Directed and Influencer-Directed Social Media Marketing on Customer Engagement and Purchase Behaviour. Australasian Marketing Journal, 33(1), 87-95. https://doi.org/10.1177/14413582241247391

This brief is written for practitioners from the published article. Open access (CC BY-NC 4.0): free to read at the publisher. Numbers are as reported by the authors; read the article for full methods.

Keep reading

More issues on Influencers · Social media · Advertising

2026 · No. 26 ·

The brand walks. The influencer burns.

In the Ferragni and Balocco case, 96% of journalists' tweets were about the influencer. Consumers followed the press, and the brand barely got mentioned.

2025 · No. 10 ·

Why a pet's sponsored post can beat a person's

Lavertu et al. find pet influencers can out-persuade people: a post with a dog's paws drew triple the engagement, and a cat lifted the price offered for wine.

Bruno Schivinski

Bruno Schivinski is Associate Professor of Marketing at Gdańsk University of Technology. He studies how consumers engage with brands in digital and social media.

About Bruno · All Marketing in 3 issues · Subscribe