Why Manual Decision-Making Is the Biggest Bottleneck in Paid Media
Human hesitation slows systems designed for speed. When decisions lag, AI optimizes around incomplete inputs and outdated assumptions. The goal isn’t removing humans—it’s placing them where judgment actually matters.
TL;DR
- Manual decision-making in paid media doesn’t just slow campaigns; it actively breaks the machine learning feedback loops that modern ad platforms depend on to function correctly.
- Google's own guidance warns that any major change to a live campaign forces a full relearning phase of 7–14 days, meaning every late or reactive decision compounds into measurable performance loss.
- West Monroe's 2026 research found that 73% of business leaders estimate their organizations lose up to 5% of annual revenue from slow decision making process failures, a "Slowness Tax" with a direct equivalent in wasted ad spend.
- By the end of 2024, 95%+ of retail advertisers had adopted AI-native campaign types like Performance Max, meaning most teams are now running machine-speed systems through human-era approval workflows, a structural mismatch that costs real money.
- The answer is not removing humans from paid advertising; it is repositioning them as system designers, governance owners, and strategic input providers rather than reactive bid adjusters.
Why Is Manual Decision-Making Costing You Money in Paid Media Right Now?
Manual decision-making is expensive in paid media because modern ad platforms are not built to wait for humans. Google, Meta, and other major platforms run on machine learning systems that require stable signals, fast conversion data, and consistent inputs, and every time a human delays a decision or makes a reactive tweak, those systems have to restart their calibration.
The result is not neutral. It is actively damaging. According to
Google's Smart Bidding documentation, smart bidding algorithms need up to 50 conversion events or three full conversion cycles to stabilize after any major change. If your team is making incremental adjustments every few days based on incomplete data, you are not optimizing; you are continuously resetting a learning process that never gets to finish.
This matters at scale.
Tinuiti's Q4 2024 Digital Ads Benchmark Report found that Performance Max accounted for approximately 69% of Shopping spend among retailers running both PMax and Standard Shopping. When the majority of your budget is inside an algorithmic campaign type, and your internal decision making process is slower than the platform's optimization cycle, you are not in control; you are in the way.
The discomfort of admitting that is exactly where growth starts. The paid advertising ecosystem has evolved to operate at machine speed, and human hesitation is now the primary bottleneck to optimal performance. Teams that fail to adapt will continue to see volatility, wasted spend, and suboptimal results, while those that restructure their decision making process will unlock the full potential of AI advertising.
What Does a Broken Feedback Loop Actually Look Like in Paid Advertising?

A broken feedback loop in paid advertising looks like volatility you cannot explain, wasted spending, and performance swings that seem to come out of nowhere. In most cases, those "mysterious" drops trace back to one cause: human intervention at the wrong moment, based on incomplete data.
Google's conversion delay documentation makes this explicit. If your average conversion cycle is 3 days and you are reviewing performance over the last 3 days, you are systematically undercounting conversions. A manager who looks at that data and pulls back spend or adjusts bids is not making an informed decision; they are making a reactive one based on a reporting artifact, and the algorithm absorbs that interference as a new signal.
This is how teams end up in a pattern that feels like bad platform performance but is actually self-inflicted turbulence:
- Incomplete data triggers human concern.
- Humans make changes to "fix" it.
- The algorithm resets its learning phase.
- Performance dips during recalibration.
- The dip triggers another human change.
- The cycle repeats.
Google's Demand Gen ramp-up guidance addresses this directly, recommending that advertisers wait approximately 2 weeks or 50 conversions before adjusting the campaign structure and limiting bid changes to ±15 % to prevent volatility. The guidance exists because the problem is common and the cost is real. For teams running paid advertising at any meaningful scale, the implication is clear: the speed and frequency of your internal decision making process is either an asset or a liability. There is no neutral.
The impact of this broken feedback loop extends beyond individual campaigns. When human decisions are made reactively, the entire paid media strategy becomes fragmented. Teams end up chasing short-term fixes rather than building long-term optimal performance. This is particularly problematic in AI advertising, where algorithms thrive on consistency and clear signals. Every unnecessary adjustment disrupts the learning process, leading to suboptimal outcomes and wasted budget.
Is the Paid Media Ecosystem Already Running at Machine Speed Without You?
Yes, and the transition happened faster than most teams realized. The paid media ecosystem has already consolidated around AI-native campaign types, which means the question is no longer whether automation will run your campaigns, but whether your human workflows are structured to work with it or against it.
Consider what the data shows:
- By the end of 2024, more than 95% of retail advertisers running Google Shopping had adopted Performance Max.
- Skai's Q2 2024 Quarterly Trends Report found that half of Skai's search accounts were using PMax and roughly one-third were using Meta Advantage Shopping Campaigns+. Automation was already mainstream by mid-2024—and it has only accelerated since.
The performance argument for AI-native campaigns is also strengthening. PMax's "sales per click" performance was approximately 1% lower than Standard Shopping in Q4 2024, compared to 13% lower in Q4 2023. The gap is closing because the machines are getting better at paid media optimization. The remaining constraint is not platform capability. It is the human infrastructure layered on top of it.
Teams still operating with weekly optimization meetings, multi-layer creative approvals, and spreadsheet-based pacing are running human-era workflows inside machine-speed systems. That mismatch has a cost, and
West Monroe's research puts a number on it: 73% of leaders say their organizations lose up to 5% of annual revenue from slow decision making process execution. In ad spend terms, that translates directly to missed seasonality windows, delayed creative rotation, slow response to cost-per-click spikes, and late promo launches.
The shift to AI-native campaigns is not just a trend; it is a fundamental change in how paid advertising operates. Teams that continue to rely on manual decision-making will find themselves increasingly outpaced by competitors who have embraced AI advertising. The key to optimal performance lies in aligning human workflows with the speed and efficiency of machine learning systems.
How Does AI Advertising Create Competitive Pressure Even Outside Your Campaigns?
AI advertising is not just changing how campaigns are managed; it is changing where high-intent consumers start their purchase journeys, which means slow internal decision making is causing teams to optimize toward a funnel that is already shifting beneath them.
Adobe Analytics data from March 2025 reported that traffic to U.S. retail sites from generative AI sources grew approximately 1,300% from November to December 2024 versus the prior year, and was up approximately 1,200% in February 2025 versus July 2024, doubling roughly every two months since September 2024.
More importantly,
Salesforce's 2025 holiday shopping data found that AI-referred shoppers converted nine times more often than social referrals. Traffic from AI search channels like ChatGPT and Perplexity doubled year over year during the 2025 holiday season.
This matters for AI advertising strategy in a direct and uncomfortable way. If your team's internal approval process takes a week, and high-intent demand is shifting to new discovery surfaces every two months, "we will decide next week" means you are running campaigns calibrated for a consumer environment that no longer exists. Caution does not preserve accuracy when the market is moving faster than your decision making process can track.
Optimal performance in this environment requires more than good creative and accurate targeting. It requires a decision infrastructure that matches the pace of the systems you are operating inside. Teams that fail to adapt will find themselves optimizing for outdated consumer behaviors, leading to wasted spend and missed opportunities. The competitive pressure from AI advertising is not just about campaign management; it is about staying ahead of shifts in consumer behavior and ensuring your paid media strategy evolves accordingly.
What Is the Real Cost of Manual Decisions Compared to AI-Assisted Systems?

The comparison between manual and AI-assisted decision systems in paid media is not abstract; it shows up in conversion rates, cost-per-lead, response times, and revenue.
Here is a direct comparison based on documented outcomes:
| Decision Type | Manual Approach | AI-Assisted Approach | Measured Outcome |
|---|---|---|---|
| Bid Optimization | Weekly review cycles, reactive adjustments | Continuous real-time calibration | PMax sales-per-click gap narrowed from 13% below Standard Shopping (Q4 2023) to 1% below (Q4 2024) |
| Lead Response Time | Multi-step human routing and qualification | Automated lead scoring and routing | -75% lead response time in 60 days (BusySeed SaaS client) |
| Cost Per Lead | Manual audience adjustments, human pacing | Automated targeting and budget allocation | -29% cost-per-lead across locations (BusySeed franchise client) |
| Revenue Growth | Campaign changes batched through approval layers | Automation-driven abandoned cart and scoring | +41% monthly revenue in 3 months (BusySeed e-commerce client) |
| Decision Speed (Org-Wide) | Consensus-based, multi-layer sign-off | Defined SLAs with automated triggers | 73% of leaders report up to 5% revenue lost from slow decisions (West Monroe, 2026) |
| Creative Volume | Human-limited iteration cycles | GenAI-assisted production | 86% of video ad buyers using or planning GenAI for creative (IAB, 2025) |
The data does not suggest that removing humans produces better results. It suggests that repositioning humans, away from reactive micro-decisions and toward system design and governance, is where optimal performance actually comes from. In paid advertising, the cost of manual decisions is not just inefficiency; it is lost revenue, missed opportunities, and a competitive disadvantage in an increasingly automated landscape.
At
BusySeed, working with 500+ businesses across more than 100 services, the consistent pattern is the same: when decision latency is reduced, and automation handles the operational layer, teams get time back for the kind of judgment that actually moves the needle. This shift is not about replacing humans; it is about empowering them to focus on strategic decisions that drive optimal performance in AI advertising.
Where Should Humans Actually Sit in an AI Advertising System?
Humans should sit at the level of system design, not campaign steering. The framing that AI advertising replaces human judgment is wrong and counterproductive, because it leads teams to either resist automation entirely or abdicate oversight in ways that create real risk.
The
IAB State of Data 2025 report found that only 30% of agencies, brands, and publishers have fully integrated AI across the campaign lifecycle. Half the industry lacks a strategic roadmap for doing so. That gap is not primarily a technology problem; it is a workflow and governance problem.
Separately,
IAB's August 2025 report on responsible AI in advertising found that more than 70% of marketers had experienced an AI-related incident, such as hallucinations, bias, or off-brand content, yet fewer than 35% planned to increase investment in AI governance and brand integrity over the next 12 months. That is a meaningful exposure for any brand running AI advertising at scale without structured human oversight.
The
NIST Generative AI Risk Management Framework, published in July 2024, provides a practical, government-backed model for structured human oversight: governance, risk management, testing protocols, and incident disclosure frameworks. It is a useful reference point for positioning human involvement as a designed system rather than an improvised approval chain.
In practical terms, for paid advertising teams, human judgment belongs in these places:
High-value human inputs:
- Conversion taxonomy and event structure
- Offline conversion imports and CRM feedback loops
- Lead quality signals and LTV segmentation
- Promotional calendars and pricing inputs
- SKU-level margin data for smart bidding targets
Human-designed guardrails:
- Target CPA and ROAS bands
- Budget pacing rules and spend caps
- Brand safety parameters and creative compliance standards
- Audience exclusion logic
What humans should stop doing:
- Making reactive bid changes based on incomplete conversion windows
- Adjusting campaign structure before the learning phases are complete
- Running weekly "optimization" passes that reset algorithmic calibration
- Approving creative one piece at a time when volume is the constraint
This is the redesign that produces optimal performance: fewer decisions about operational details, more decisions about the inputs and constraints that shape what the system optimizes toward. In AI advertising, humans are not obsolete; they are essential for defining the rules, setting the goals, and ensuring the system operates within the bounds of brand safety and business objectives. The key is to reposition human involvement from reactive adjustments to proactive governance.
How Do You Restructure the Decision Making Process for Speed Without Losing Control?
You restructure the decision making process by replacing ad-hoc reviews with deliberate decision SLAs, response tiers that define what gets decided immediately, what waits 48–72 hours, and what is evaluated weekly. Speed becomes intentional rather than chaotic, and the algorithm gets the stability it needs to reach optimal performance.
Google's guidance on learning phases and
smart bidding calibration provides a natural framework for designing those tiers. When you know that a change triggers a 7–14 day learning phase and requires 50 conversions to recalibrate, you build your internal rhythm around that reality instead of fighting it.
Here is a structured checklist for restructuring the decision making process in a paid media operation:
How Do You Build a Decision Redesign Checklist for Paid Media Teams?
The fastest path to improving paid media outcomes is not finding better tools; it is designing clearer decision rules. This checklist gives your team a working framework.
Decision Redesign Checklist for Paid Media Teams:
- Audit your current decision latency. Map every recurring decision in your campaign management workflow, bids, budgets, creative approvals, audience changes, and document how long each actually takes from identification to execution.
- Categorize decisions by urgency tier. Same-day decisions should cover broken tracking, disapprovals, budget caps hit, and landing page outages. 48–72 hour decisions cover creative fatigue interventions and audience adjustments. Weekly decisions cover portfolio budget shifts, channel mix reviews, and experiment readouts.
- Document your conversion delay by funnel stage. Pull your "days to conversion" data per campaign type and create a rule that delays performance judgments by that window. Do not evaluate the last three days of a campaign with a three-day conversion cycle.
- Batch structural campaign changes into scheduled release windows. Treat campaign changes the way product teams treat software releases, deliberate, documented, and spaced to allow evaluation before the next round of changes.
- Apply the 50-conversion rule before judging new campaigns. Per Google's smart bidding guidance, wait for at least 50 conversions or three conversion cycles before making structural changes to a new campaign or a recently modified one.
- Cap reactive bid changes at ±15%. For AI-native campaign types like Demand Gen, Google explicitly recommends limiting bid adjustments to ±15% to avoid triggering volatility and resetting learning.
- Define human inputs into the system, not constant steering of it. Assign team owners for conversion taxonomy, offline conversion imports, LTV segmentation, and promotional calendar inputs; these are the levers humans should be pulling.
- Establish AI governance checkpoints, not approval bottlenecks. Build a review cadence for brand safety, creative compliance, and measurement integrity, but separate those governance checkpoints from the operational campaign decisions that need to move fast.
- Tie creative approvals to volume thresholds, not individual pieces. When 86% of video ad buyers are using or planning GenAI for creative production, the bottleneck is approval speed, not creative capacity. Design approval workflows around batches, not one-by-one review.
- Review this framework quarterly, not annually. The consumer environment and platform capabilities are changing so quickly that AI-referred traffic to retail sites doubled every two months from September 2024 onward, so a decision framework designed for last year's environment may already be optimizing for a funnel that has moved.
Restructuring the decision making process is not about sacrificing control; it is about gaining control over the factors that truly drive optimal performance in paid advertising. By aligning human workflows with the speed and efficiency of machine learning systems, teams can reduce volatility, improve results, and unlock the full potential of AI advertising. The key is to design a system where automation handles the operational details, and humans focus on strategic inputs and governance.
What Does Optimal Performance Actually Require From Human Teams in 2025 and Beyond?
Optimal performance in modern paid media requires human teams to stop competing with automation for control over operational details and start owning the inputs, constraints, and governance structures that enable automation to work correctly. The goal is not fewer humans; it is better-positioned humans.
The practical proof is in the outcomes.
BusySeed's SeedTech automation framework increased monthly revenue by 41% for an e-commerce apparel client within 3 months, driven by lead scoring and abandoned-cart automation. A SaaS client saw a 75% reduction in lead response time and a 38% increase in demo bookings within 60 days through multi-channel automation. A national franchise group reduced cost-per-lead by 29% across locations. In each case, the win came from removing decision latency at the operational layer, not from removing human judgment from the strategic one.
AI advertising tools are doing the same work at the platform level. When the algorithm handles bid calibration, audience expansion, and placement decisions in real time, the human team's role is to define what the algorithm is optimizing toward and ensure the inputs it receives are accurate and current. That means clean conversion tracking, reliable offline conversion imports, margin data that reflects real business economics, and promotional signals that arrive before the promotion launches—not after.
The teams that will see sustained optimal performance from their paid advertising investments are not the ones with the most sophisticated manual oversight. They are the ones who have designed a system where automation handles speed and volume, humans handle truth and direction, and the two operate in a rhythm that does not require either to compensate for the other's weaknesses. This balance is the future of paid media, and it requires a fundamental shift in how teams approach the decision making process.
At
BusySeed, we have spent years building that kind of infrastructure for clients across more than 100 service categories. The pattern holds: when decision latency is designed out of the operational layer, human judgment gets sharper because it is applied where it actually has leverage. The result is not just better-paid media performance; it is a more efficient, more effective, and more scalable approach to AI advertising.
FAQ
Q1) What are the best AI decision-making tools for small businesses running paid advertising on limited budgets?
The best AI advertising tools for small businesses are the ones already embedded in the platforms you are using, not separate software purchases. Google's Smart Bidding, Meta's Advantage+ Shopping Campaigns, and Microsoft's automated bidding strategies are all machine learning systems designed to optimize paid advertising outcomes without requiring manual bid management. These tools are particularly effective for small businesses because they leverage vast amounts of data to make real-time adjustments that a human would be unable to replicate.
The challenge for small businesses is not access to these tools; it is understanding how to feed them accurately. For a small team, the most valuable investment is not a new tool but a cleaner decision making process around conversion inputs. That means making sure your conversion tracking is correctly configured, your conversion window matches your actual sales cycle, and your offline conversion data (if applicable) is being imported reliably. A small business running Smart Bidding with clean data and a stable campaign structure will consistently outperform one making frequent manual adjustments based on incomplete reporting.
For businesses that need support building that infrastructure, partners like
BusySeed can provide the expertise and tools needed to maximize optimal performance in paid media. The key is to focus on the inputs that drive the system, rather than trying to control every operational detail.
Q2) What are the best AI tools for predictive business analytics in paid media campaigns?
The best AI tools for predictive business analytics in paid media are those that integrate seamlessly with your existing platforms and provide actionable insights. Google's Performance Planner and Meta's Advantage+ Shopping Campaigns are excellent starting points, as they use machine learning to forecast performance and suggest optimizations. These tools are particularly valuable because they leverage platform-specific data to make highly relevant predictions for your campaigns.
For more advanced predictive analytics, tools like Skai and Marin Software offer cross-platform insights and predictive modeling capabilities. These tools can help you anticipate trends, identify opportunities, and make data-driven decisions that drive optimal performance. However, the effectiveness of these tools depends on the quality of the data they receive. Ensuring clean, accurate, and comprehensive data inputs is critical for maximizing the value of predictive analytics in AI advertising.
Small businesses can also benefit from tools like Google Analytics 4, which provides predictive metrics such as purchase and churn probabilities. These metrics can help you identify high-value audiences and tailor your paid advertising strategy accordingly. The key is to use these tools to inform your decision making process, rather than relying on them to make decisions for you. Human judgment is still essential for interpreting the data and applying it to your specific business context.
Q3) What are the top solutions for integrating AI in human decision-making for paid media?
The top solutions for integrating AI in human decision-making for paid media focus on augmenting human judgment rather than replacing it. Tools like Google's Smart Bidding and Meta's Advantage+ Shopping Campaigns are designed to handle the operational details of campaign management, freeing up human teams to focus on strategic inputs and governance. These tools use machine learning to make real-time adjustments to bids, budgets, and targeting, allowing humans to define the goals and constraints that guide the system.
For more advanced integration, platforms like Skai and Marin Software offer AI-driven insights and recommendations that can inform human decision-making. These tools provide predictive analytics, performance forecasts, and optimization suggestions, helping teams make more informed decisions about their paid advertising strategy. The key is to use these tools to enhance the decision making process, rather than relying on them to make decisions autonomously.
Another critical solution is the use of AI-powered creative tools, such as those offered by Canva and Adobe. These tools can generate high-quality creative assets at scale, allowing human teams to focus on strategic creative direction rather than production. This integration of AI in creative workflows can significantly improve the efficiency and effectiveness of AI advertising campaigns.
Ultimately, the best solutions for integrating AI in human decision-making are those that align with your specific business needs and goals. The key is to design a system in which AI handles operational details, and humans focus on strategic inputs and governance. This balance is essential for achieving optimal performance in paid media.
Q4) How can small businesses leverage AI advertising to compete with larger competitors?
Small businesses can leverage AI advertising to compete with larger competitors by focusing on the strengths of AI-driven tools: speed, efficiency, and data-driven decision-making. AI-native campaign types like Google's Performance Max and Meta's Advantage+ Shopping Campaigns are designed to optimize performance in real time, allowing small businesses to achieve results that would be impossible with manual management. These tools level the playing field by automating the operational details of campaign management, freeing up small teams to focus on strategic inputs and creative direction.
The key to competing with larger competitors is to design a decision making process that aligns with the speed and efficiency of AI-driven tools. This means reducing decision latency, ensuring clean and accurate data inputs, and focusing on strategic governance rather than reactive adjustments. Small businesses that embrace this approach can achieve optimal performance in paid media without the need for large teams or extensive resources.
Another advantage of AI advertising for small businesses is the ability to scale quickly. AI-driven tools can handle large volumes of data and make real-time adjustments, allowing small businesses to expand their reach and compete in markets that would otherwise be inaccessible. By leveraging the power of AI, small businesses can achieve results that rival those of larger competitors, while maintaining the agility and flexibility that are often their greatest strengths.
For small businesses looking to maximize the impact of paid advertising, partners like
BusySeed can provide the expertise and tools needed to design and implement an effective AI-driven strategy. The key is to focus on the inputs that drive the system, rather than trying to control every operational detail.
Q5) What are the key considerations for implementing AI tools in the decision-making process for paid media?
The key considerations for implementing AI tools in the decision making process for paid media revolve around alignment, governance, and data quality. First, it is essential to align the AI tools with your business goals and objectives. This means defining clear goals for your campaigns, such as target CPA, ROAS, or conversion volume, and ensuring that the AI tools are configured to optimize toward those goals. Without this alignment, AI-driven tools may make adjustments that do not align with your business priorities.
Second, governance is critical to ensuring that AI tools operate within your brand safety and compliance standards. This means establishing clear guidelines for creative compliance, audience targeting, and brand safety, and ensuring that AI tools adhere to them. Human oversight is essential for monitoring compliance and making necessary adjustments. The
NIST Generative AI Risk Management Framework provides a useful model for designing governance structures that ensure responsible AI use in AI advertising.
Third, data quality is a foundational consideration for implementing AI tools in paid media. AI-driven tools rely on accurate, comprehensive, and timely data to make informed decisions. Ensuring clean conversion tracking, reliable offline conversion imports, and accurate margin data is essential for maximizing the effectiveness of AI tools. Without high-quality data, AI-driven tools may make suboptimal decisions, harming performance.
Finally, it is important to design a decision making process that balances the speed and efficiency of AI with the strategic judgment of humans. This means defining clear roles for humans and AI, ensuring that humans focus on strategic inputs and governance, while AI handles the operational details. This balance is essential for achieving optimal performance in paid advertising.
Works Cited
- Adobe. "Adobe Analytics: Traffic to U.S. Retail Websites from Generative AI Sources Jumps 1200 Percent." Adobe Blog, 17 Mar. 2025, blog.adobe.com/en/publish/2025/03/17/adobe-analytics-traffic-to-us-retail-websites-from-generative-ai-sources-jumps-1200-percent.
- BusySeed. "How BusySeed's SeedTech Drives Revenue Growth Through Automation." BusySeed, www.busyseed.com/how-busyseeds-seedtech-drives-revenue-growth-through-automation.
- Google. "About Conversion Delay." Google Ads Help, support.google.com/google-ads/answer/12262960.
- Google. "About Smart Bidding." Google Ads Help, support.google.com/google-ads/answer/13020501.
- Google. "Demand Gen Ramp-Up Guidance." Google Ads Help, support.google.com/google-ads/answer/15110871.
- IAB. "AI Adoption Is Surging in Advertising—But Is the Industry Prepared for Responsible AI?" IAB, 21 Aug. 2025, www.iab.com/insights/ai-adoption-is-surging-in-advertising-but-is-the-industry-prepared-for-responsible-ai/.
- IAB. "IAB State of Data Report 2025." IAB, www.iab.com/news/iab-state-of-data-report-2025/.
- IAB. "Nearly 90% of Advertisers Will Use GenAI to Build Video Ads." IAB, 15 July 2025, www.iab.com/news/nearly-90-of-advertisers-will-use-gen-ai-to-build-video-ads.
- NIST. "Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence." NIST, July 2024, www.nist.gov/publications/artificial-intelligence-risk-management-framework-generative-artificial-intelligence.
- Salesforce. "2025 Holiday Shopping Data." Salesforce News, 8 Jan. 2026, www.salesforce.com/news/stories/2025-holiday-shopping-data/.
- Skai. "Key Takeaways from the Skai Q2 2024 Quarterly Trends Report." Skai Blog, skai.io/blog/key-takeaways-from-the-skai-q2-24-quarterly-trends-report/.
- Tinuiti. "Digital Ads Benchmark Report Q4 2024." Tinuiti Research Insights, tinuiti.com/research-insights/research/digital-ads-benchmark-report-q4-2024/.
- West Monroe. "Slow Decisions Are Costing Companies Millions." West Monroe Press Releases, 27 Jan. 2026, www.westmonroe.com/press-releases/slow-decisions-are-costing-companies-millions.
