The Customer Problem

When you open Spotify, the homepage is different for every user. When you browse Nike's site, products are ranked by what Nike thinks you'll buy. When you log into Chase, offers are tailored to your spending. This is web and mobile personalization: deciding in real time what each customer sees inside your product. It directly moves conversion, retention, and revenue at the point where customers are already engaged.

Almost nobody does it well outside of companies that build it from scratch. A single experiment (testing whether high-spend users convert better seeing a loyalty upsell on the account page) requires an analyst to build the segment, a PM to write targeting rules, a designer to create variants, an engineer to instrument the test, and someone to monitor results. Most teams run 5–10 per quarter and end up with a handful of static rules covering a fraction of their users.

The tools that exist (Optimizely, Adobe Target) handle infrastructure but don't solve the coordination problem. Every experiment, audience, and piece of content still requires manual configuration. They also force companies to duplicate customer data into a proprietary store, adding months of implementation and permanent vendor lock-in.

Why Agents are the Solution

The shift from "AI assists a step" to "AI owns a workflow" is already happening in software engineering. Personalization is ready for the same shift.

Today: A growth team testing whether high-intent visitors convert better with a comparison page spends 2–4 weeks coordinating across tools and people. They run maybe 8 experiments that quarter.

With agents: The growth lead describes the goal. An agent identifies the audience from existing customer data, generates variants informed by past performance, configures the experiment with holdout groups, launches it, monitors significance, and reallocates traffic. The human reviews before launch. Calendar time: hours. The same team runs 50+ experiments per quarter because the coordination cost drops to nearly zero.

Agents handle the full chain as one workflow: audience, content, experiment, optimization. Instead of a bunch of manual steps, agents rethink this as a continuously improving closed loop.

Why Hightouch, Why Now

More than 1,000 enterprises—like Domino’s, Chime, Spotify, Ramp, Whoop, Grammarly, use Hightouch to run their marketing. Earlier this year we launched the Agentic Marketing Platform with two pillars:

• Agentic Advertising Studio. Agents analyze performance, competitor creative, and brand guidelines, then assemble ad concepts across Meta, Google, TikTok, and LinkedIn.
• Agentic Lifecycle Studio. Agents identify lifecycle opportunities, draft audiences, assemble on-brand email and SMS campaigns, and orchestrate them through Braze, Iterable, Salesforce, and Adobe.

Advertising covers paid media. Lifecycle covers email, SMS, and push. Agentic Personalization is the third pillar: what customers experience inside the website and mobile app. Different buyer (product and growth teams), different workflow (real-time experimentation), different technical surface (browser SDKs and edge delivery). Same agent infrastructure and data foundation.

What no competitor can replicate:

• Agent orchestration. 25+ specialized skills for audience building, content generation, campaign analysis, and multi-step workflow coordination.
• Content assembly. Agents search existing asset libraries for reusable on-brand content before generating anything new, which is what makes output trustworthy enough for enterprises to ship.
• Marketing and brand context. Persistent layer connecting customer data, brand guidelines, creative assets, and performance history so agents operate grounded in how the business works.
• Real-time data infrastructure. Low-latency personalization API, sub-second audience evaluation, ML-powered optimization and recommendations, event pipeline processing millions of events daily.

The Role

Founding Product Manager for Agentic Personalization: Hightouch's third major pillar and biggest new investment. You define the product, the sequencing, and the go-to-market.

• Customer discovery. Deep engagement with product and growth teams at enterprise companies to understand how they personalize web and mobile today, and what would earn their trust in an agent-driven approach. Translate into a roadmap tied to revenue milestones.
• Define the agent experience. What agents do autonomously (segments, variants, experiments, traffic allocation) vs. what needs human approval (production launches, brand-sensitive content). Design the feedback loop that builds trust over time.
• Scale the GTM. Co-own pipeline, join customer calls, shape competitive positioning, partner with CS on implementation playbooks.

High-visibility role working directly with founders and executive team.

What We're Looking For

• Zero-to-one track record. Products you built from nothing to meaningful revenue, ideally at a B2B SaaS company. Domain matters less than evidence you can find product-market fit under uncertainty.
• Technical depth. Can evaluate architectural tradeoffs, debate infrastructure decisions with senior engineers, and reason about things like batch vs. real-time and caching strategies at the level of product decisions.
• Commercial instinct. Think in terms of pipeline, conversion, and expansion. Build roadmaps that map to revenue targets.
• Speed. Small team, high autonomy, tight cycles. We use AI heavily in our own workflows.
• Cross-functional trust. This spans engineering, design, sales, CS, and leadership. You drive clarity from ambiguity.

We are looking for talented, intellectually curious, and motivated individuals who are interested in tackling the problems above. This is a senior role, but we focus on impact and potential for growth more than years of experience. The salary range for this position is $240,000 - $320,000 USD per year, which is location independent in accordance with our remote-first policy. We also offer meaningful equity compensation in the form of ISO options, and offer early exercise and a 10 year post-termination exercise window.

The Problem

Every large advertiser runs the same loop: strategist reviews performance data, decides what to test next, briefs a designer, waits for assets, hands them to a media buyer who builds campaigns across Meta and Google and TikTok, launches, monitors, repeats. The process is slow because creative production, campaign management, and performance analysis happen in three separate systems, and insights travel between them through decks, Slack threads, and weekly syncs.

Companies like Nike, Uber, and PepsiCo spend millions per month on digital ads and employ entire teams just to keep this loop turning. Most teams run 10 to 15 creative tests per month because every test requires manual work across multiple tools and teams. The result: most ad spend runs against stale creative, performance degrades from fatigue, and the gap between "what's working" and "what gets made next" is measured in weeks.

Why Agents Change This

Advertising is uniquely high-leverage for AI agents because nearly every step in the workflow is now automatable, and the steps compound:

• Research. Agents can continuously monitor competitor ad libraries, auto-classify creative by format, hook, and message, and surface positioning gaps. What used to require a strategist spending hours in Meta Ad Library happens in the background, always current.
• Creative production. Generative models can now produce ad-quality images and video. The constraint has shifted from "can AI make this" to "can AI make this on-brand and on-spec for each platform." That is a solvable engineering problem, not a research problem.
• Campaign structure. Building properly structured campaigns (naming conventions, audience splits, placement targeting, budget allocation) across five ad platforms is tedious, rules-based work. Agents can do it faster and more consistently than humans.
• Optimization. Detecting creative fatigue, reallocating budget toward winners, and knowing when to rotate in fresh variants is pattern recognition across structured data. Agents excel at exactly this.
• The compounding effect. When one agent handles the full chain (research, production, activation, analysis, iteration) as a single workflow, the cycle time collapses from weeks to hours. A team that runs 15 tests per month can run 50+, not because any individual step got faster, but because the coordination cost between steps drops to nearly zero.

Today: A performance marketing team testing new concepts on Meta and TikTok spends 2 to 3 weeks per cycle.

With agents: The marketing lead describes the objective. An agent analyzes recent performance, identifies the highest-value concepts to test, generates on-brand variants across formats, builds campaigns, launches after human approval, monitors results, detects fatigue, and queues replacements. Same team, 5x to 7x the testing volume, because the human bottleneck becomes strategy and judgment, not execution.

Why Hightouch, Why Now

Hightouch built the Agentic Marketing Platform: one system where AI agents own entire marketing workflows across lifecycle (email, SMS, push through Braze, Iterable, Salesforce, Adobe), advertising (creative production and activation across paid channels), and product experiences (real-time web and mobile experimentation). All three surfaces share the same agent infrastructure, brand and customer context, and warehouse-native data foundation. An insight the ads agent learns about creative performance can inform what the lifecycle agent sends. That shared context across every marketing surface is the product.

Ads is already live, and demand is surging. Enterprise customers are generating ad creative today with agents that analyze performance across Meta, Google, TikTok, LinkedIn, and Snapchat, produce briefs grounded in competitive research and brand guidelines, and generate ready-to-ship ad concepts. Early adoption has been strong: customers who start using it quickly expand usage across campaigns and teams, and inbound interest from enterprise prospects is accelerating. The foundation works. The opportunity is building it into a platform that captures the full advertising workflow.

What Hightouch has that no competitor can replicate:

• Agent orchestration. 25+ specialized agent skills for audience building, content generation, campaign analysis, and multi-step workflow coordination, all composable into end-to-end workflows.
• Content assembly with brand trust. Agents search existing asset libraries for reusable on-brand content before generating anything new. This is what makes output trustworthy enough for enterprises to ship without heavy review cycles.
• Marketing and brand context. A persistent layer connecting customer data, brand guidelines, creative assets, competitive intelligence, and performance history so agents operate with full business context, not generic prompts.
• Warehouse-native data foundation. 62,000 workspaces, a million data syncs per day, 300+ integrations. Customer data stays in the warehouse. No proprietary data store, no duplication, no lock-in.

The Role

You will lead the Agentic Ads Platform. The foundation exists (customers generating ads today), and the opportunity is turning it into the platform that closes the full advertising loop. Four major areas of work:

• Multi-model creative generation. Today the platform generates static images. The roadmap includes video generation, motion graphics, format adaptation, and multi-model orchestration (routing to the best model for each task). Building a creative generation pipeline that produces broadcast-quality output across formats is a hard, high-value problem.
• Marketing context and brand adherence. Making agents genuinely trustworthy for enterprise creative teams. Better context engineering so agents follow brand guidelines, business rules, legal requirements, and channel-specific policies without constant human correction. The gap between "AI-generated ad" and "ad our brand team would actually approve" needs to shrink dramatically.
• Platform depth across ad channels. Deep integrations with Meta, Google, TikTok, LinkedIn, Snapchat, and others. Not just launching ads, but understanding each platform's creative specs, optimization levers, and best practices so agents can automatically adapt creative for each channel, syndicate content across platforms, and optimize placement-level performance. Think: an agent that knows a 9:16 video with text-safe zones works differently on TikTok Reels vs. Instagram Stories and adjusts accordingly.
• Best-in-class agentic creation experience. The interaction model for how marketers work with agents to go from objective to live campaign. This means the conversational interface, the review and approval workflows, the feedback loops that help agents learn what "good" looks like for each brand, and the progressive autonomy controls that let teams start with full human approval and graduate to agent-led execution as trust builds.

Day to day, you will:

• Run deep customer discovery with performance marketing and creative teams at enterprise companies. Understand their workflows, pain points, and what would earn their trust.
• Define the product roadmap tied to revenue milestones. Sequence what to build, what to compose from existing systems, and what to partner on.
• Ship product. Work directly with engineering and design on tight cycles.
• Co-own pipeline and go-to-market. Join customer calls, shape competitive positioning, partner with CS on implementation.

High-visibility role working directly with founders and executive team.

What We're Looking For

• Zero-to-one track record. You have built products from nothing to meaningful revenue, ideally at a B2B SaaS company. The domain matters less than evidence you can find product-market fit under uncertainty.
• Technical depth. You can evaluate architectural tradeoffs, reason about things like multi-model orchestration and API integration design at the level of product decisions, and hold your own in debates with senior engineers.
• Commercial instinct. You think in terms of pipeline, conversion, and expansion. You build roadmaps that map to revenue targets.
• Speed. Small team, high autonomy, tight cycles. We use AI heavily in our own workflows and expect you to do the same.
• Cross-functional trust. This role spans engineering, design, sales, CS, and leadership. You drive clarity from ambiguity.