IAB Australia today published Future Search Skills, a 21-page paper from its Future of Search Working Group that names three skills search marketers will need over the next few years and argues that automation is removing the routine work through which junior practitioners used to learn the trade. The paper draws on the experience of working group members from brands, agencies, publishers, platforms and retail media networks, on discussion at the industry body's Discovery: AI & Search Summit in July, and on a survey of 12 specialist search and performance marketing recruiters conducted in October 2026.
The recruiter numbers are small, and the paper says so. On future demand for search roles the recruiters are split, but on skills and junior hiring they point the same way. According to IAB Australia, 58% of the recruiters said demand for search and performance roles had fallen over the past 12 to 24 months, 92% said demand for AI-related skills had risen, and two thirds expect demand for junior roles to keep falling over the next two to three years.
In Short
IAB Australia's search working group today published a paper naming three skills search marketers will need: judging what AI systems recommend, working across paid, organic, AI and social search, and pairing search depth with a second specialty. It matters because platforms now automate bidding, campaign setup and reporting, and a small survey of 12 recruiters in the paper found two thirds expect demand for junior roles to keep falling. What changes is how people learn the trade, with the paper asking employers, platforms and universities to rebuild entry-level training around diagnosis, testing and judgement instead of manual execution.
A paper written by the people it describes
The Future of Search Working Group is made up of practitioners, not academics, and three of its 18 listed members work for platforms whose products appear in the paper as examples: Google, Microsoft and Reddit. Agencies including WPP Media, OMD, Spark Foundry and iProspect sit alongside publishers such as News Corp Australia and The Guardian, advertisers such as Mars and Suncorp, and Cartology, the Woolworths Group retail media business.
According to the paper, its aim is "to help existing search professionals, in both SEM and SEO, expand their skills, and to help people preparing to enter the workforce understand how search-related roles are changing and what they will need to succeed in them."
| Member | Organisation |
|---|---|
| Dan Aguirre-Davies | WPP Media |
| Jarrod Allen | Monks |
| Meg Bellemore | The Guardian |
| Greg Cattelain | Spark Foundry |
| Mike Cook | News Corp Australia |
| Bonnie Dodemaide | iProspect |
| Simon Dunwoody | Woolworths Group |
| Clare Farrugia | Hatched |
| Ben Froggatt | Cartology |
| James Hutchinson | Ryvalmedia |
| Diana Kaldelis | Suncorp |
| Jules Lau | Mars |
| Rhett Mitchelhill | Half Dome |
| Max Prosser | OMD |
| Renee Stopps | Microsoft |
| Tom Tilney | |
| Anne Tran | Admatic |
| Adrian Vallelonga |
The same group has published on search before. In June 2025 it released research on social search, which PPC Land covered as social media becoming a search engine for younger shoppers, and in June 2026 it produced a 14-page paper on the agentic AI shift for search marketers. The new paper turns from how consumers search to the people who run search campaigns.
Search is still the largest line in the budget
"Search is not shrinking," the paper states. Search advertising in Australia reached A$8.6 billion in the 2026 financial year and remains the largest digital advertising investment sector, according to IAB Australia. Total internet advertising reached A$19.8 billion, the strongest growth in four years. When those figures came out at the end of August, PPC Land reported that search held a 43% share of the market and grew 13.2% on the previous year.
The paper also cites Google handling more than 5 trillion searches a year, and notes that people who research on platforms like ChatGPT "still tend to return to search to find the brand and buy." What has changed, in the paper's framing, is where people search and the fact that AI agents increasingly do the searching for them. Discovery now runs across search engines, AI answers, social and video platforms, retail marketplaces and AI assistants.
One agency search lead at the July summit put it in a line the paper repeats: most teams are still built to run search as a channel, "when it has become a behaviour."
That shift moves practitioner work upstream, toward planning, stakeholder management, the evaluation of campaigns and creative, and closer contact with chief financial officers. The paper insists the work still needs someone who understands how automated systems operate, a role it describes as human in the loop. One recruiter quoted in the paper went further:
"Platform proficiency matters more than ever, even at senior level. People used to move off the tools as they progressed. That's no longer advisable."
Five eras, from manual bids to agents
The paper divides the history of search platforms into five "activation eras," ending in an agentic period in which agents connect advertising platforms and data through standards such as MCP, the Model Context Protocol. The paper's glossary defines it as "an open standard that lets AI tools and agents connect to other software and data, such as advertising platforms." The dates and descriptions below are the paper's.
| Era | Period | Focus | Impact on the practitioner |
|---|---|---|---|
| 01 Manual Search | 2000 - 2015 | Manual bidding, keyword match types, manual quality score management, hand-built ad copy testing | "Practitioner value = technical execution." Managing the levers was where the expertise sat |
| 02 Smart Push | 2015 - 2020 | Smart Bidding, Responsive Search Ads launched | Value shifts from pulling levers to feeding good inputs and signals |
| 03 Consolidation | 2021 - 2025 | Performance Max launch, Enhanced Conversions launched | Push towards a consolidated campaign approach and stronger measurement signals |
| 04 Generative Search | 2025 - present | AI Mode, AI Overviews | A new wave of search with AI-powered placements |
| 05 Agentic | 2026 onward | Agents act for consumers (research, compare, buy) and for marketers (connecting platforms and data through protocols such as MCP) | Automation of search campaigns end to end, with practitioners setting the goals and guardrails |
The fifth era is the one the paper is written for: practitioners set goals and guardrails while agents act on both sides of the auction, for the consumer who searches and for the marketer who buys.
Skill one: judgement over AI-driven search
The first skill is AI-driven search, though the paper is careful about what that means. Using AI, it argues, is "becoming a baseline expectation rather than a differentiator on its own." The skill is judgement: questioning what automated systems recommend, checking it against the business objective, and knowing when a person needs to make the call.
The paper illustrates the point with a familiar case. An automated campaign finds more conversions, lowers cost per acquisition or uncovers new query demand. On paper, the optimisation is working. But the extra leads may be lower quality, the growth may come from a market the business cannot service efficiently, or the campaign may be optimising toward form submissions when the business cares about qualified opportunities and revenue. "These are not necessarily optimisation problems. They are business judgement problems," the paper says.
AI also opens analysis that rarely had time before. One agency search lead described matching search postcodes to socioeconomic data from the Australian Bureau of Statistics to check whether campaigns were reaching the right people, and to compare who searches with who clicks through.
The paper cites Microsoft's account of the marketing services group Stagwell, which reduced some paid search campaign audits "from hours to minutes" using a Copilot agent connected to Microsoft Advertising data. It adds a qualification: "Microsoft's other published MCP examples still retain human reviewers before changes go live." In June, Microsoft moved its Advertising MCP server into an open pilot with read-only access to live campaign data, naming Stagwell among the pilot partners.
Andrew Ng, founder of DeepLearning.AI, is cited for what he called a human "context advantage": people carry years of customer, organisational and situational context that may never reach the AI system. The paper turns this into a list of questions. Has the market changed? Has the competitive position changed? Is the proposition still compelling? Are the right conversion and customer signals being fed back into bidding? Is search being asked to solve the right acquisition problem? The question it wants raised most often is this one:
"Is this improving the business outcome and strategic direction of the campaign, or mainly improving the metric the platform has been asked to optimise towards?"
Clients are also changing. "With AI giving clients easy access to information, they question decisions more, so the ability to influence and back up the work matters more," one recruiter said.
The governance gap
Tools are getting easier to use; the operating model around them is not. The paper cites Deloitte's 2026 State of AI in the Enterprise research for Australia: 61% of Australian organisations report improved efficiency from AI, while only 30% say they are using AI to deeply transform how they work. Deloitte also reports that 69% are using autonomous AI agents, while only 22% have advanced agent governance models.
"More automation does not necessarily mean a more mature system," the paper says. Maturity, in its definition, is whether an organisation has deliberately decided what should be automated and to what extent, what still needs practitioner review, what business and customer data can be used, who owns the decision when the system recommends an action, how risk and brand controls are managed, and what success is measured against. Flagging search-term anomalies or summarising an account audit may be relatively low-risk; reallocating major budget, expanding into new demand, changing brand messaging or optimising against sensitive customer data may need far more human involvement.
The paper points to two IAB Australia resources: its AI-Ready Advertising Infrastructure, covering organisational maturity, data governance, vendor assessment, model risk, generative AI usage policies and incident response, and its 2026 LLM Prompting Framework. In November 2025, IAB Australia's AI Working Group published an LLM prompting guide built around context, role, objective and task.
How the learning framework works
According to IAB Australia, the framework for building practical AI capability in search has four stages, running from core capability to greater technical complexity. The paper calls Apply and Govern "the centre of gravity for most practitioners, with Govern as the key focus." The stages, with the paper's examples, are:
- Understand. Build enough AI literacy to use it safely and to know where it appears across search planning and management. The examples: follow platform documentation, AI usage guidelines, workshops and industry updates; treat outputs as drafts and check claims, numbers and sources before use; understand privacy, security and approved-use requirements before sharing business or customer data; and know whether AI is "assisting, recommending, optimising or acting."
- Apply. Use AI features intentionally, monitor results closely and learn through testing. In automated bidding, the platform manages auction-time bids "while the practitioner still defines the conversion goal, value signal and efficiency target." For Google's AI Max for Search, the paper's wording is "Test expanded matching, asset customisation and URL expansion with the right controls in place." AI-assisted analysis is validated against the business question, and experiments or A/B tests come before scaling.
- Govern (key focus). Measurement quality, business-value signals, attribution, quality assurance, privacy, risk and human review. The examples: optimise to qualified leads, not form fills, by feeding CRM outcomes back into bidding; measure commercial outcomes such as ROI alongside the effect of the AI feature itself; check AI-generated assets against brand, legal and category requirements; challenge recommendations against the market, customer and objective; and capture learnings for future decisions.
- Orchestrate. Connected workflows, agents and automation "where they solve a real business problem." The paper's example is Stagwell's Assembly, which used a Copilot agent connected through Microsoft Advertising MCP to cut some audits from hours to minutes, "with practitioners reviewing outputs and making the final call." Connected workflows use APIs, MCPs or agents "with clear permissions, approval points and ownership."
AI Max carries its own trade-offs. When Google released its implementation guide in 2025, PPC Land reported that enabling Final URL expansion means pinned headlines are not respected, the kind of control question the Apply stage is meant to surface.
Skill two: whole-of-search visibility
The second skill responds to convergence. Where PPC and SEO teams once competed in silos for separate positions on the results page, AI search results put both in competition for the same space. The paper also widens the definition of search. Nearly half of people online aged 16 to 34 say searching for products is one of their main reasons for using social media, and Reddit reaches more than 17 million Australians a month, according to IAB Australia's 2025 social search research. "Whole-of-search visibility has to include them," the paper says.
There is a division of labour inside that convergence. SEO and GEO work focuses on content such as buying guides and comparisons, while paid search handles more pointed searches for product or service details. The paper's glossary defines generative engine optimisation as "the practice of improving how a brand and its content appear, are cited and are described in AI-generated answers. It builds on SEO."
The paper cites BrightEdge research finding that half of ChatGPT's holiday citations point to pages marketing does not own, and that commercial pages are the most cited content type in Google AI Overviews, pages that typically belong to merchandising or marketing teams rather than to a GEO strategy.
It also borrows a model from Search Engine Journal's State of PPC 2026 report: a move from the "T-shaped" marketer, one deep specialism supported by broad knowledge, to an "M-shaped" one with two or three areas of deep expertise. The paper quotes the description, "It looks like a generalist from a distance and like a specialist up close," and the report's list of skills now expected of a competent PPC manager: "data analysis, first-party data activation, creative testing strategy, attribution modeling, prompt engineering, and scripting."
"I expect greater demand for professionals who can connect search, performance media, data and commercial outcomes, rather than operate within a single platform or channel," one recruiter said.
Quality Score 2.0
Quality Score, in the paper's account, created a "relevancy economy" in which the most relevant sponsored results win the top positions, not simply the highest bids. The open question is how AI-powered search will change authority, relevancy, ranking and Quality Score itself.
The evidence the paper assembles is mixed. It cites a Search Engine Journal report headlined on AI Overviews cutting organic clicks by 38%, and notes the rise of zero-click searches. Yet it also says recent IAB Australia research on commerce and discovery found click-through from search "is holding and time on search is up 22%." Sponsored results inside AI-powered search grew 394% between March and October 2025, with 25.6% of AI searches including sponsored results, according to Semrush data the paper cites.
PPC Land has tracked similar measurements. A randomized field experiment by researchers at the Indian School of Business and Carnegie Mellon University found outbound organic clicks fell 39.8% when an AI Overview appeared, and Ahrefs put the click-through decline for top-ranked pages at 58% in February. A Semrush study published in October found AI Overviews on 88% of the commercial keywords it tracked in August, and in May PPC Land documented a sponsored product listing inside an AI Mode answer, marked only by a small "Sponsored" label.
On copy, the paper admits the industry is "caught in the limbo." Clients are asking whether LLM marketing principles should be applied to search, and particularly to sponsored placements in AI environments. Ad relevance is one of three Quality Score components, but it can also affect expected click-through rate. The paper's conclusion is that "platform growth and business impact are not linear (yet)."
A blind spot in the reporting
Paid search teams have long explained week-to-week movement with auction insights, Google Analytics and market tools on competitor behaviour. AI search results create "a growing blindspot in reporting and optimisation, which can't be tracked through traditional search platforms and reports," according to the paper.
It cites a BrightEdge finding that winning in AI Mode depends on three inputs that sit in different teams in a traditional agency.
| AI Mode optimisation input | Agency team that usually owns it |
|---|---|
| Product feed quality | PPC |
| Schema | SEO/GEO |
| On-page content | SEO/GEO |
AI visibility and tracking reports usually sit with SEO and GEO teams, yet paid search marketers are now being asked for the same insight. Some agencies are reorganising: teams covering organic, paid, retail and emerging search are briefed together and, in the most advanced cases, share KPIs, "so success is judged on total search rather than each channel defending its own numbers." The paper floats one possible unified metric, a "Total Search Share/Presence" combining paid visibility, organic visibility, AI visibility, brand demand and attention metrics. It is offered as an example, not a defined measure.
Measuring total search
Traffic used to stand in for success, and the paper says that proxy is weakening. Practitioners are replacing it with engaged sessions, micro-conversions such as product page views, citations and sentiment in AI answers, and engagement on social, read alongside paid and organic results.
The new metrics carry a risk. Most AI visibility tools rely on simulated prompts rather than real user data, so their outputs are directional, and platforms use different names for similar measures. At the July summit, an audience member asked where the evidence was that these metrics convert to revenue. The panel's answer, as the paper records it, was that measurement has not kept pace and that many are, "for now, vanity metrics."
The paper lists four steps it calls a core skill: start from a measurement framework and the business questions it must answer, not from a tool; hold one consistent source of business truth, such as sales, qualified leads or revenue; treat AI visibility data as directional and test whether it moves that truth; and use incrementality testing and marketing mix modelling to connect paid and organic search activity to outcomes.
"ROI alone is no longer enough. Demand is growing for experience with CLV, contribution margin, MER, POAS, marginal CPA and incrementality. That relies on client data that hasn't always been on offer to agencies, so the experience is rare, and an opportunity as more clients open up," one recruiter said.
The paper's glossary defines the measurement terms as follows.
| Term | Definition in the paper |
|---|---|
| CLV | Customer lifetime value. "The total value a customer is expected to bring a business over the whole relationship." |
| MER | Marketing efficiency ratio. "Total revenue divided by total marketing spend, across all channels." |
| POAS | Profit on ad spend. "Profit generated for each dollar of advertising spend, rather than revenue." |
| Marginal CPA | "The cost of each additional conversion as spend increases." |
| MMM | Marketing mix modelling. "Statistical method that estimates the contribution of each marketing channel to sales or other outcomes." |
For journeys run by agents, the paper refers back to the working group's June paper and its four tests: was the brand considered, did the interaction earn trust, did the agent complete the task, and did it lead to the next step. PPC Land's coverage of that paper described the same structure as a four-stage framework of consideration, confidence, completion and continuation.
Brand becomes a search input
When AI builds the shortlist, the paper argues, the brands most likely to make it are the ones people already know and trust. Each category draws on different sources for AI answers, from Reddit and reviews to creators and affiliates, and search marketers need to know which ones shape answers in their category. "Pushing more budget into capturing existing demand is not a long-term plan if nothing is building the demand in the first place," the paper says. In September, IAB Australia's creator research found 61% of Australian online shoppers often or sometimes discover brands through creators.
Skill three: a hybrid skill stack
"Search built its authority on technical complexity," the third section begins. Bidding structures, match types, Quality Score and shopping feeds gave specialists a defendable skill set for years; automation has eroded that moat. Search depth still matters, the paper insists, but the practitioners with the most leverage pair it with a second capability that solves a broader business problem.
The default pairing in media agencies, paid search with paid social, is under pressure. Paid social is becoming a creative-led discipline in its own right, and platform fragmentation makes running both to a high standard harder for one person. For publishers, the paper says organic search expertise has to move closer to editorial and content skills, developing what it calls "GEO-to-Editorial" discernment.
"You either find people who are good in-platform and with data, or people who are good at thinking outside the box, strategy and client leadership. It's very difficult to find both, and we need both," one recruiter said. Another described the most valuable search and performance marketers of the next two years as "those who can direct and audit AI-run campaigns, feed algorithms better conversion and profit signals, and win visibility in AI search tools like ChatGPT and Google's AI Overviews."
The paper sets out six personas that share core search skills but add different specialties.
| Persona | Combines | Why it works, per the paper | Who it is for, per the paper |
|---|---|---|---|
| Performance | Search + Paid Social | Owns the full paid performance budget, so spend moves to whatever channel is working instead of being defended by channel | Someone equally comfortable with a spreadsheet and a creative brief |
| Signals | Search + Measurement & Tracking | Checks the inputs feeding the automation and "catches bad data before it becomes a bad decision" | Someone drawn to data who wants to see behind the platform's curtain |
| Discovery | Paid Search + SEO/GEO + Social Search | Defends visibility across paid, organic, social and AI-led discovery, "not just their half" | Someone who thinks about how a brand is found everywhere |
| Builder | Search + AI Workflow Design | Builds and maintains the agents and workflows connecting tools and data, "not just use them" | Someone curious about how systems connect, who likes automating repetitive work |
| Commerce | Search + Retail Media | Retail marketplace placements run on the same auction mechanics as search; as AI agents shop for consumers, feed quality becomes a deciding factor | Someone commercially minded with a taste for feeds and product data |
| Full-Funnel | Search + Video/Upper-Funnel | Connects search intent and upper-funnel buying instead of running separate budgets | Someone who wants to be a "full funnel marketer" beyond performance specialism |
"The shift isn't toward one universal search marketer who can do everything. It's toward the right combination for the problem in front of you," the paper says.
What 12 recruiters said
The working group surveyed 12 specialist digital and performance marketing recruiters in October 2026. "The sample is small but experienced, so the results are directional rather than representative," the paper says. Because each percentage rests on 12 responses, PPC Land has converted them into approximate headcounts.
| Finding | Share reported | Recruiters out of 12 (PPC Land calculation) |
|---|---|---|
| Demand for search and performance roles fell over the past 12 to 24 months | 58% | 7 |
| Expect demand for those roles to grow over the next two to three years | Half | 6 |
| Expect demand for those roles to fall over the next two to three years | 42% | 5 |
| Have seen demand for junior roles fall | 58% | 7 |
| Have seen demand for junior roles fall significantly | A third | 4 |
| Expect junior demand to keep falling over the next two to three years | Two thirds | 8 |
| Demand for AI-related skills has risen | 92% | 11 |
| Most sought AI skill: optimising content for AI search experiences | 83% | 10 |
| Most sought AI skill: automating search tasks and workflows | 83% | 10 |
| AI-assisted research and analysis | 75% | 9 |
| Data analysis using AI tools | 75% | 9 |
The most common hiring challenge was finding people with strong hands-on platform skills who can also think strategically, understand the commercial picture and lead clients. Some recruiters also find it hard to tell platform knowledge from proven results. Recruiters expect demand to grow for incrementality, marketing mix modelling, customer lifetime value and profit-based metrics, alongside AI search optimisation, creative testing, first-party data and retail media.
The pipeline is thin. Recruiters pointed to a small local talent pool, universities teaching little performance marketing and strong competition for good candidates, which pushes up salaries and counter-offers. Three quotes from the survey sum up the mood:
"Fewer roles, higher bar. One strong person can now cover what a small team used to, so each hire needs to be hands-on, commercial and able to build."
"Fewer junior roles now means a thinner mid-level in a few years."
"Universities in Australia are not teaching enough performance marketing to graduates coming out of marketing degrees, so the dependence on overseas candidates and visa holders is high across the market."
The recruiter view matches IAB Australia's own hiring data. The paper cites the 2026 Digital Advertising & Ad Tech Industry Talent Review, which found entry-level roles made up as little as 1% of open vacancies, while around half of advertised roles asked for six or more years' experience. PPC Land's report on that review in June put the experienced share at 49%, from data gathered across 54 organisations.
The junior paradox
The paper's central argument fits in one line: "we can automate the task, but we can't automate the learning."
For years, it says, search advertising provided "an almost accidental apprenticeship." Junior staff learned by building campaigns, testing ad copy, adjusting bids, interrogating search terms, pulling reports and working out why performance moved. Much of that work was manual and repetitive, and technology is rightly making it faster. But it also taught people how advertising works.
"Gaps that would once have been seen as a development opportunity are now seen as a step too far," one recruiter said.
The paper does not argue for keeping manual tasks as a rite of passage. Nor does it accept that someone new can "leap straight into strategy." Its answer is to redefine the basics around understanding the system: setting the right objective, providing the right inputs, questioning recommendations, interrogating outputs, understanding consumer behaviour and connecting media performance to commercial outcomes. Juniors, in this model, become "an essential human in the loop," with apprenticeships built around experimentation, diagnosis and judgement instead of repetitive execution.
It asks whether the final year of university could become more vocational, whether students could work on real industry briefs, and whether internship and graduate pathways could combine hands-on work with mentoring and exposure to AI-powered advertising. "We don't need to protect the old tasks. We need to protect the learning those tasks created," the paper says. It closes the section with a question for every practitioner: "What do I want my human advantage to be?"
The skills checklist
The paper ends with a nine-row checklist for current practitioners and people moving into search, each rated "Strong," "Building" or "Not yet." The entries below follow the paper's wording, condensed.
| Skill | What good looks like | Starting point for those already in search | Starting point for those new to search |
|---|---|---|---|
| Platform fundamentals | Understands how the auction, Quality Score, bidding and product feeds work, even when they are automated | Stay hands-on in the platforms with seniority; check automated recommendations before accepting them | Learn how the auction and Quality Score work; complete Google and Microsoft Advertising certifications |
| AI judgement | Questions what AI systems recommend and checks it against the business objective before acting | For each automated change, ask why it is happening, what context is missing and whether it improves the business outcome | Use AI tools on real tasks and check every output; follow IAB Australia's LLM Prompting Framework |
| Commercial understanding | Connects search activity to revenue, margin and customer value, and can explain it to finance teams | Learn the metrics finance teams use: CLV, contribution margin, MER and marginal CPA | Learn how a business makes money: margin, acquisition cost and the value of a customer over time |
| Measurement | Starts from a measurement framework, uses incrementality testing and MMM, and treats AI visibility data as directional | Run an incrementality test on part of the activity; learn how MMM results feed into search budgets | Learn analytics basics and the difference between correlation and incremental impact |
| Whole-of-search | Sees paid, SEO, GEO, social search, retail media and AI answers as one system | Learn the basics of the adjacent discipline; track how brands appear in AI answers | Search the category on Google, ChatGPT, TikTok and Reddit and compare the results |
| Signals and data | Feeds automation the right conversion, customer and profit signals, and catches bad data early | Review conversion tracking and check whether bidding optimises to qualified leads and value, not form fills | Learn how tracking and conversion data flow into automated bidding |
| Building with AI | Builds and maintains workflows and agents that connect tools and data, with clear controls | Build one workflow that saves the team time, document it and share it | Learn basic scripting or a no-code automation tool |
| Communication and influence | Explains and defends recommendations in business terms, not platform metrics | Present performance as business outcomes; prepare for clients who arrive with their own AI-generated analysis | Present analysis to the team and seek feedback |
| Hybrid depth | Pairs search depth with one adjacent specialty that solves a broader business problem | Choose a persona and plan the training and experience to get there | Try more than one area early before choosing where to build depth |
The call to industry
The paper closes by asking five groups to replace deliberately the training that automation is removing.
| Group | What the paper asks |
|---|---|
| Employers (agencies, brands and publishers) | Keep hiring at entry level; redesign junior roles around diagnosis, testing and judgement; protect time for learning; give juniors supervised access to live platforms and pair them with experienced practitioners |
| Platforms | Extend training beyond feature set-up to scenario-based problem solving; give students and new starters sandbox access to AI-powered campaign tools |
| Universities | Work with industry on final-year projects using real briefs, data and platforms |
| Practitioners | Choose the adjacent skill to build depth in, and start now |
| IAB Australia and the Future of Search Working Group | "Commitments to university engagement, training resources, and industry updates" |
Why this matters for the marketing community
The paper lands at an awkward moment for search teams. Spend is rising, with search taking the largest share of the A$19.8 billion Australian internet advertising market, while demand for the people who run it, particularly at entry level, is falling, according to the recruiters surveyed. IAB Australia's talent review in June showed how few junior openings remain; the recruiter survey published today suggests the people who fill roles do not expect that to reverse soon.
The skills it describes are a response to platform changes PPC Land has covered throughout 2026. AI Overviews appeared on 88% of the commercial keywords Semrush tracked in August, sponsored listings sit inside AI Mode answers, and Microsoft has opened read-only access to its advertising data for AI agents through an MCP pilot. Each shifts work from operating campaigns to checking what automated systems do, which is the skill the paper lists first.
There is also a structural point. The paper says that in an AI Mode environment, product feeds, schema and on-page content belong to different agency teams, and that the most advanced agencies now share KPIs across paid, organic and retail search. That is an organisational question as much as a training one, and it lands on agencies at the same time as the pressure to "deliver more with less" the paper describes in its introduction.
Readers will weigh the source. The working group includes representatives of Google, Microsoft and Reddit, and two of the paper's practical examples, AI Max for Search and the Stagwell audit case study, come from those platforms' own products and published materials. The recruiter survey, by the paper's own account, is directional.
What the paper does not say
The paper does not name the 12 recruiters or their firms, publish the questionnaire, or split the findings between paid search and SEO roles. It gives no salary figures. The research behind the "time on search is up 22%" figure is described only as recent IAB Australia work on commerce and discovery. The "Total Search Share/Presence" metric has no formula, and IAB Australia's own commitments carry no dates or targets.
Key facts
- Document: Future Search Skills, subtitled "skills to future-proof search marketing"; web page title "Future Search Skills: Preparing for Search Marketing"
- Publisher: IAB Australia, Future of Search Working Group (18 listed members)
- Publication date: November 19, 2026
- Format: 21-page PDF, 4 MB
- Recruiter survey: "Recruiter perspectives on the changing skills and roles in search and performance marketing," 12 specialist recruiters, October 2026; described by the paper as "directional rather than representative"
- Other inputs: working group expertise; IAB Australia Discovery: AI & Search Summit, Agencies & Brands panel, July 2026
- Three skills: AI-driven search; whole-of-search visibility; building a hybrid skill stack
- Learning framework stages: Understand, Apply, Govern (key focus), Orchestrate
- Personas: Performance, Signals, Discovery, Builder, Commerce, Full-Funnel
- Market figures cited: search advertising A$8.6 billion in FY26; internet advertising A$19.8 billion in FY26
- Third-party sources cited: BrightEdge, Search Engine Journal (State of PPC 2026), Semrush, Microsoft Advertising (Stagwell case study), Deloitte (State of AI in the Enterprise: Australia, 2026), OAIC guidance on commercially available AI products (2024)
- Next related event: IAB Australia Discovery: AI & Search Summit Sydney 2027, scheduled for March 16, 2027
- Web page: https://www.iabaustralia.com.au/resource/future-search-skills/
- PDF: https://www.iabaustralia.com.au/wp-content/uploads/2026/11/IAB-Australia-Future-Search-Skills-2026.pdf
Timeline
- 2000 - 2015: The paper's "Manual Search" era: manual bidding, match types and hand-built ad copy testing
- 2015 - 2020: "Smart Push" era: Smart Bidding and Responsive Search Ads launch
- 2021 - 2025: "Consolidation" era: Performance Max and Enhanced Conversions launch
- May 2025: Google releases its AI Max for Search campaigns implementation guide
- June 2025: IAB Australia's Future of Search Working Group publishes social search research showing nearly half of people aged 16 to 34 search for products on social media
- March - October 2025: Sponsored results in AI-powered search grow 394%, according to Semrush data cited in the paper
- November 19, 2025: IAB Australia's AI Working Group releases an LLM prompting guide
- February 2026: Ahrefs finds AI Overviews correlate with a 58% click-through decline for top-ranked pages
- May 25, 2026: A sponsored product listing appears inside a Google AI Mode answer, marked only by a small "Sponsored" label
- June 2, 2026: IAB Australia's talent review finds entry-level roles at 1% of open vacancies
- June 2026: The Future of Search Working Group publishes its paper on agentic AI for search marketers
- June 17, 2026: Microsoft moves its Advertising MCP server into an open, read-only pilot, with Stagwell among partners
- July 2026: IAB Australia's Discovery: AI & Search Summit, where panellists call many AI visibility metrics "vanity metrics" for now
- August 31, 2026: IAB Australia reports A$19.8 billion in FY26 internet advertising, with search at A$8.6 billion
- September 2026: IAB Australia creator research finds 61% of online shoppers discover brands through creators
- October 2026: The working group surveys 12 specialist search and performance recruiters
- October 21, 2026: Semrush finds AI Overviews on 88% of tracked commercial keywords
- November 19, 2026: IAB Australia publishes Future Search Skills
- March 16, 2027: IAB Australia Discovery: AI & Search Summit Sydney 2027 scheduled
Related PPC Land coverage
- Australia's ad industry is running out of junior talent and it's not coming back - IAB Australia's 2026 talent review, with entry-level roles at 1% of open positions.
- IAB Australia maps the agentic AI shift search marketers can't ignore - The same working group's June paper on how AI agents compress the search journey.
- Social media becomes new search engine for younger shoppers - The working group's 2025 research on product search inside social platforms.
- Australian internet ad market hits $19.8 billion in FY26, up 14% - The FY26 revenue report in which search reached A$8.6 billion.
- IAB Australia publishes LLM prompting guide for marketing professionals - The AI Working Group's vendor-agnostic prompting framework from November 2025.
- Microsoft bets on AI economy with Web IQ, Clarity citations and MCP server - Microsoft's read-only Advertising MCP pilot, with Stagwell among the partners.
- Google releases AI Max for Search campaigns implementation guide - Google's guide to the feature the paper uses as its Apply-stage example.
- Researchers find Google AI Overviews cut publisher clicks 39.8% - A randomized field experiment on organic clicks when AI Overviews appear.
- Google's AI summaries now swallow 58% of clicks that once went to websites - Ahrefs' analysis of 300,000 keywords and top-ranked page click-through.
- Google AI Overviews now hit 88% of tracked commercial keywords, Semrush finds - AI Overview coverage across banking, software and ecommerce keywords.
- Google AI Mode ads blend into answers and a billion users are watching - How sponsored listings appear inside AI Mode responses.
- IAB Australia: 85% of under-40s find brands through creators - IAB Australia's creator research on brand discovery.
Summary
Who: IAB Australia, through its Future of Search Working Group of 18 practitioners from agencies, brands, publishers, retail media and platforms including Google, Microsoft and Reddit, with input from 12 specialist recruiters.
What: A 21-page paper, Future Search Skills, naming three skills search marketers will need (AI-driven search, whole-of-search visibility and a hybrid skill stack), a four-stage AI learning framework, six hybrid personas, a nine-row skills checklist and a call to employers, platforms, universities and practitioners. Its recruiter survey found 92% reporting higher demand for AI skills and two thirds expecting junior demand to keep falling.
When: Published November 19, 2026. The recruiter survey ran in October 2026, and the paper draws on the July 2026 Discovery: AI & Search Summit.
Where: Australia, where search advertising reached A$8.6 billion in FY26, though the platform changes it describes, from AI Overviews to agent connections through MCP, are global.
Why: Automation now handles much of the bidding, setup and reporting that once trained junior staff, entry-level roles have fallen to as little as 1% of advertised vacancies, and the paper argues the industry has to rebuild that learning deliberately while practitioners shift from operating platforms to judging what they do.
Discussion