Beyond bots and quick fixes: The future of advertising platform support

Transforming ad support through AI-driven solutions and early issue detection to enhance customer trust and efficiency

Why the old support model is dead

Many online advertising platforms face rising complexity, skyrocketing support volumes, and increasingly frustrated customers — from small advertisers and creators to major brands and agencies.

But here’s the truth: most platforms are struggling with outdated service models. They lean on adding more agents, leveraging chatbots that only deflect, and layering in new tools without fixing the core problem, enabling a better experience. The support experience is not designed for the customer; it’s designed for the company.

The result? Bloated help centers full of jargon; Long wait times for human help; Opaque, trust-eroding decisions; and revenue leakage and rising churn.

We believe it’s time for a full reinvention — not incremental tweaks but a fundamental redesign — of how online ad platforms deliver customer support. In this article, we look at six of the most common problems with current service models and propose the best ways ad platforms can tackle them.

What often goes wrong and how to get it right 

01

Design help centers and taxonomies for the customer

What goes wrong: Help centers and supporting taxonomies are written for internal classifications, not customer questions.
Real scenario: A small advertiser gets a policy violation and can’t figure out what they did wrong. But the help center displays irrelevant rules, not solutions.
How to get it right: Build customer-centered flows carrying solutions directly relevant to their problems using generative artificial intelligence (GenAI)-powered search and guided resolution that understands intent, not just keywords. Taxonomies should be simple, direct, and reflect the voice of the customer. The design of the help center needs to be based on structured data in the advertiser context and run on a system blending GenAI with iterative search, actions, and escalations.

02

Replace dead-end automation with a system powered by AI agents

What goes wrong: Chatbots and canned replies acknowledge problems but don’t solve them.
Real scenario: A major ad campaign is stuck in review; the bot says, “we’re working on it,” but no action is taken for days, costing the customer revenue or potentially limiting their ability to spend money with you.
How to get it right: Use agentic automation that can take multistep actions (resubmit, escalate, pause spend) without human intervention. To bring this to life, AI agents must not only be trained on advertiser-specific data and actions, but should also have multi-step reasoning built on user intent data (not keywords). The agents will need to focus on the entire spectrum of automated actions rather than just specific use cases, while human-in-the-loop systems should be included for safety and learning.
 

03

Make systems find your customers before they need help

What goes wrong: Systems wait for customers to discover problems, even when internal data shows clear warning signals.
Real scenario: A travel company’s payment method is expiring, but no alert is sent and campaigns shut down mid-flight.
How to get it right: Build proactive service triggers designed around the customer journey with early notifications and guide customers through immediate fixes. For planned outages, preemptively communicate to customers utilizing AI agents for service mitigation. Develop early warning systems by aggregating signals from past support tickets and resolutions to proactively detect and prevent catastrophic-level issues before they escalate.

04

Measure prevention, not just speed

What goes wrong: Teams optimize response time and closure rates but ignore why customers needed help in the first place. Thus, larger pain points never get addressed.
Real scenario: An advertiser repeatedly contacts support about the same disapproval issue; it’s resolved each time, but the root cause is never fixed.
How to get it right: Connect engineering beyond native product experience. Extend product design to eliminate unnecessary support call. Instead of just handling it faster, shift to root-cause analytics and taxonomies that reduce inbound demand at the source.

05

Get ruthless about transparency

What goes wrong: Customers are given vague policy violations or black-box decisions, with no clarity or progress updates.
Real scenario: A creator's account is suspended or compromised, and they want to recover their content; they appeal but hear nothing for weeks, fueling online backlash.
How to get it right: Use plain-language explainers and transparent appeal tracking to build trust and reduce frustration. Embed this into your help center so that it is easily accessible.

 06

Don’t just solve — teach

What goes wrong: Service fixes individual issues but misses the opportunity to educate and upskill customers, especially in situations where errors are user-driven. It also misses the opportunity for customers to help one another with crowd-sourced solutions.
Real scenario: A small business repeatedly complains about poor ad performance but never receives optimization guidance and is left confused and frustrated.
How to get it right: Pair resolutions with personalized, actionable insights that improve future outcomes. Connect customers so they can share solutions and best practices. Reward on how your customer rates your experience and not just operational metrics.

Key design elements for CX and UX in customer support and help centers

To make these changes stick, platforms must embed world-class customer experience (CX) and user experience (UX) principles into every layer of the help experience:

1

Plain language, not jargon

Simplify content, responses, and policy explanations into clear, human language. During onboarding, provide users with actionable steps, checkboxes and a progress bar to reduce uncertainty for users.

2

Embedded help, not just external links

Place help directly inside key workflows e.g., ads managers, billing dashboards, and creative tools.

3

Recognition, not recall

Make objects, actions, and options visible to reduce memory load, striking a balance by combining images, icons, and video tutorials, reducing cognitive load and improving readability. Feature an autocomplete search function that predicts relevant articles and then presents a summarized solution generated by AI.

4

Proactive notifications with clear actions

Send warnings and include immediate action buttons or guided fixes. Keep users informed about what is happening with clear feedback.

5

Clear visual hierarchy

Use design signals (color, layout, progress bars) to show where customers are in the journey. Include clickable breadcrumbs at the top of each article, improving navigation by indicating users’ current location within the help center. Provide shortcuts for experienced users.

6

Personalized recommendations

Surface relevant next steps or resources, driven by customer behavior and context. Incorporate a diagnostic survey that helps identify user challenges and offers tailored, scenario-based solutions. Include notes throughout help articles, offering best practices, tips, and case studies. This helps users avoid common errors and learn from real-world examples.

7

Customer feedback

Allow users to submit feedback on specific sections of an article as well as on their general experience with solutioning. This approach ensures more targeted and effective content updates.

8

Mobile-first experiences

You don’t win customer loyalty by answering more tickets you win by eliminating the need for tickets in the first place. You don’t win by adding more agents you win by making your systems smarter, your processes clearer, and your human talent more strategic. The platforms that rethink support now will own the next decade of customer trust, growth, and market leadership.

How KPMG can help

Instead of just throwing more agents or bots at the problem, KPMG helps you challenge the entire system by:

  • Treating customer support as a product design challenge, not an operational patch.
  • Focusing on automation that resolves, not just deflects.
  • Helping teams reduce support demand by fixing underlying issues, not scaling busy work.
  • Designing human roles to maximize strategic impact, not fill gaps left by broken systems.
  • Embedding trust and transparency into every process, making the help center a competitive differentiator.

We help reinvent your programs for online advertising support through:

1

Help center diagnostic and redesign — Assess friction points, broken flows, and content gaps.

2

GenAI and agentic automation blueprint — Identify where and how to apply smart automation that actually drives resolution.

3

Proactive service layer design — Set up predictive models and alerts to catch issues before they hit.

4

Human workforce reengineering — Redesign agent roles, tools, and escalations to focus human talent on what matters.

5

Transparency and trust overhaul — Rewrite policies, decision communications, and appeal systems for clarity and fairness.

6

Outcome-driven metrics reset — Define and track KPIs that reflect true customer and business outcomes.

Explore our insights

Meet our team

KPMG can help you modernize the way you acquire, retain, and grow customer relationships. Combining market-leading industry experience with functional acumen and technical know-how, we can help you deliver greater returns from your customer-focused investments.

Image of Aditya Rath
Aditya Rath
Principal, Customer Advisory, KPMG US
Image of Manish Shah
Manish Shah
Managing Director, Customer Advisory, KPMG US

Thank you

Thank you for subscribing. You should receive a confirmation e-mail soon.

Connect with KPMG

Opt-in to receive the latest government insights from KPMG

Thank you for reaching out. We will be in touch soon to follow up.

By submitting, you agree that KPMG LLP may process any personal information you provide pursuant to KPMG LLP's . Privacy Statement

An error occurred. Please contact customer support.

Explore other services tailored to your business​

Thank you!

Thank you for contacting KPMG. We will respond to you as soon as possible.

Contact KPMG

Use this form to submit general inquiries to KPMG. We will respond to you as soon as possible.

By submitting, you agree that KPMG LLP may process any personal information you provide pursuant to KPMG LLP's . Privacy Statement

An error occurred. Please contact customer support.

Job seekers

Visit our careers section or search our jobs database.

Submit RFP

Use the RFP submission form to detail the services KPMG can help assist you with.

Office locations

International hotline

You can confidentially report concerns to the KPMG International hotline

Press contacts

Do you need to speak with our Press Office? Here's how to get in touch.

Headline