I've started using Google Ads. I've got an App campaign that uses target Return On Ad…
- User
I've started using Google Ads. I've got an App campaign that uses target Return On Ad Spend (tROAS) bidding, and I've started with a low target to allow the algorithm to experiment a bit more. It's only been a week but results look good, I've averaged considerably above my target already and the acquired users are still providing value, so I'll end up with an even better return.
One thing confuses me though. In the first 4 days, the campaign spent all of my budget each day and has managed to beat my tROAS. But in the last 3 days, it's spent ~20% of my budget each day. It's still getting a great return, and in some cases even better than before, but it feels to me like it's not meeting its potential. It should easily be able to spend the full daily budget and reach the tROAS, it's got clear evidence of doing this previously.
There's a lot of moving parts here, but what do you think? Any experience to draw upon?
Assistant
This is a very common and understandable concern with tROAS campaigns! What you're seeing is likely a sign of the algorithm working correctly, but it requires a shift in perspective about how tROAS bidding fundamentally operates. Here's a breakdown of what's probably happening and what to consider:
Why tROAS Campaigns Reduce Spend When Performing Well:
- The Core Goal is Efficiency, Not Spending: The algorithm's primary objective is not to spend your budget. Its goal is to achieve your specified tROAS. If it can achieve that target (or exceed it) while spending less than your daily budget, it absolutely will. This is efficient behavior.
- Finding the Sweet Spot: In the initial "learning phase" (your first 4 days), the algorithm explores a wider range of auctions and users to gather data. This often requires spending the full budget. Once it identifies the most efficient users (those most likely to convert at or above your target ROAS), it focuses bids there. These users might be:
- Fewer in number than the broader audience initially targeted.
- Found in auctions with lower average costs (CPAs).
- Converting at a significantly higher ROAS than your target.
- Auction Dynamics & Competition: The landscape changes daily. Competitors' bids, user behavior, and ad inventory availability fluctuate. What was achievable spending 20/day on Day 6 because the algorithm found cheaper, high-converting opportunities or competition for those prime users decreased.
- Pacing: Google's algorithms often try to smooth spending throughout the day and avoid exhausting the budget too early, potentially missing out on later high-value auctions. While this doesn't usually lead to days of significant underspend, it can be a factor in daily fluctuations.
- Value Attribution Window: Crucially, the algorithm optimizes based on the predicted value within your conversion window. If your "value" (like in-app purchases) happens days after install, the algorithm might be seeing strong signals that the users acquired in the last 3 days are predicted to generate excellent future value, allowing it to achieve the tROAS target with fewer new installs/spend. This is why performance looks good even with lower spend.
- Diminishing Returns: There might only be a finite number of users available right now who are predicted to convert at or above your target ROAS. Spending more might force the algorithm to target users predicted to convert at a lower ROAS than your target, which it is programmed not to do.
Why This Might Be Good (and Why Patience is Key):
- Proven Efficiency: Your campaign is demonstrably efficient. Exceeding your tROAS target consistently is the primary goal of this strategy.
- Algorithm Confidence: The algorithm is showing it has learned and can hit the target with less "effort" (spend). This is a positive sign.
- Profit Protection: It's protecting your profitability by not overspending just to hit a budget cap when the efficient opportunities aren't there at scale.
What to Do & Consider (Based on Your Goal):
- Do Nothing (Monitor Closely): This is often the best initial approach, especially since it's only been a week and performance is good. Give it more time (at least 1-2 more full weeks). The algorithm is likely still learning, especially about the long-term value of users. Underspend days can be normal fluctuations.
- Gradually Increase Your tROAS Target (The Recommended Next Step): This is the primary lever for tROAS campaigns to spend more.
- Why? By raising your target (e.g., from 100% to 120%), you tell the algorithm: "Only show my ads to users predicted to be even more valuable." To find enough of these higher-value users to maintain spend volume, the algorithm will need to bid more aggressively and potentially enter more auctions, often leading to increased spend while maintaining or improving efficiency.
- How? Make small increments (e.g., 10-20% increases). Monitor closely for 3-7 days after each change. Don't jump too high too fast.
- Goal: Find the tROAS target where the campaign consistently spends most of your budget while still meeting or exceeding that higher target. This maximizes profitable scale.
- Increase Budget (Use with Caution):
- Potential Upside: If there truly is untapped, efficient inventory, increasing the budget might allow the algorithm to find more users at or near your current tROAS target.
- Significant Risk: Increasing budget without raising the tROAS target often signals the algorithm to target wider or less efficient audiences to spend the extra money. This frequently leads to ROAS dropping towards your target (or below). Only do this if you are willing to potentially accept a lower average ROAS for more total volume/profit. Generally, adjusting tROAS is preferred over adjusting budget for scaling.
- Check for Technical Issues (Less Likely, But Verify):
- Ad Approval: Ensure all ads are approved and running.
- Targeting: No accidental targeting restrictions changed?
- Billing: Payment method is valid and not causing delivery issues.
- Reporting Delay: Ensure the "under-spend" isn't just a reporting delay (especially for in-app events). Check real-time spend vs. reported spend.
- Audience Saturation: While less common for broad app campaigns, could a core audience be temporarily exhausted? (Unlikely after just a week).
- Evaluate Incrementality (Advanced): Is the campaign really acquiring users who wouldn't have installed organically? Sometimes efficient campaigns slow down because they've captured the most readily convertible users. Tools like Google Ads' incrementality experiments can help assess this later.
Key Takeaways & Experience:
- tROAS != Budget Spend: Budget is a cap, not a goal. The goal is the ROAS target.
- Underspend with High ROAS is Usually Efficient: It means the algorithm found the best users cheaply. Celebrate this efficiency!
- Scaling Requires Higher tROAS Targets: To spend more profitably, you generally need to increase your tROAS target incrementally. This pushes the algorithm to find more valuable users.
- Patience is Critical: One week is very early, especially with post-install value. Significant learning and optimization happen over 2-4 weeks. Resist the urge to make drastic changes based on a few days.
- Focus on Overall Profit, Not Just Spend or ROAS: Calculate your total profit (Revenue from Ads - Ad Spend). If profit is increasing or stable while ROAS is high, even with lower spend, it might be optimal for now.
Recommendation for Now:
- Wait: Monitor for at least another 5-7 days. See if the pattern stabilizes or if spend naturally increases again.
- Analyze: Look at the overall performance for the full week and the projected value of recent users. Is total profit acceptable/growing?
- Small tROAS Increase (After Monitoring): If after 2 full weeks, spend is consistently low (e.g., <50% of budget) while ROAS is consistently well above target (e.g., 150%+), make your first move a small increase to your tROAS target (e.g., 10-20% higher). Observe the impact on spend and ROAS for another 3-7 days.
This phased, data-driven approach based on understanding the algorithm's efficiency goals is the best way to scale your successful campaign profitably. Good luck!
- User
You suggest increasing the budget. In fact, Google Ads offers this personalised suggestion too. But if my daily budget isn't even being consumed, how does this help? Google Ads suggests that increasing my budget by £400 a week could net me an extra £150 in return, but that's below even my fairly modest tROAS and doesn't look like a good deal. It's also far, far below the rate I've been getting so it seems unlikely to be accurate as I don't believe it would drop off that quick with a bigger net.
Assistant
You've hit on a critical insight - and your skepticism about Google's budget increase suggestion is completely justified in this scenario. Here's why:
Why Google Suggests Increasing Budget (Even When Underspending)
- Algorithmic Blind Spot: The suggestion engine often focuses on budget utilization as a primary signal for "opportunity." It sees unused budget capacity and assumes you want to spend more, regardless of efficiency.
- Simplistic Projection Model: The "£400 for £150" projection is likely based on:
- Extrapolating Current CPA/ROAS: It takes your current average cost per acquisition (CPA) or value/conv and naively applies it to more spend.
- Ignoring Diminishing Returns & Target Constraints: It doesn't adequately factor in that the easiest, cheapest, highest-value users are likely already acquired. To spend significantly more, the algorithm would likely have to target users predicted to convert at a lower ROAS than your current target or even your tROAS setting.
- Ignoring Your tROAS Constraint: Crucially, the projection often doesn't strictly enforce your tROAS target in its simulation. It assumes you're willing to accept some drop in efficiency to get more volume. This is where it fundamentally misaligns with your goal (maximizing profit within your efficiency target).
Why This Suggestion is Likely Flawed & Potentially Harmful in Your Case:
- Your Evidence is Clear: You know that spending your current budget consistently yielded results well above your tROAS target. Google's projection of spending £400 more for only £150 more return implies a ROAS of just 37.5% (£150 / £400). That's massively below your current performance and your tROAS target. This huge discrepancy proves their projection model is making pessimistic (or simply inaccurate) assumptions about the availability of efficient inventory at scale.
- tROAS Bidding's Core Rule: The algorithm is programmed not to spend money if it predicts the user won't meet your target ROAS. Forcing more spend by increasing the budget cap, without changing the target, often forces the algorithm to violate its own optimization rule. It has to target less efficient users to spend the extra money, causing ROAS to plummet towards (or below) your target.
- The "Drop Off" is Real: Your intuition ("it wouldn't drop off that quick") is partially correct for the current target, but the algorithm's projection is likely assuming that to find £400 worth of EXTRA efficient users daily, it would need to dig much deeper into lower-quality segments. This is where diminishing returns hit hard. The first £20 spent daily might yield 500% ROAS, the next £80 might yield 200%, and the next £400 might only yield 37.5% – because the best users are finite.
What Google's Suggestion Should Say (But Doesn't):
"Your campaign is consistently exceeding its tROAS target while spending only 20% of its daily budget. This indicates the algorithm has found a highly efficient, but limited, pool of users. To spend more profitably while maintaining or improving efficiency, you should gradually increase your tROAS target, not your budget."
The Strategic Path Forward (Stick to Your Guns):
- Ignore the "Increase Budget" Suggestion (For Now): Based on your results and the flawed projection, increasing the budget now is highly likely to:
- Decrease your average ROAS significantly.
- Potentially lower your overall profit despite higher spend (if the return on the extra spend is poor).
- Disrupt the algorithm's current efficient optimization.
- Focus on Increasing Your tROAS Target (The Correct Lever): This is the only reliable way to tell the algorithm: "Find me more valuable users, and I authorize you to bid higher to get them."
- How Much? Start small. Increase your tROAS target by 10-20%. (e.g., if target was 100%, set it to 110% or 120%).
- Why Small? You want to nudge it into finding slightly more valuable users without causing a drastic reduction in available volume. Big jumps can shock the system.
- What to Expect:
- Goal: Spend should increase as the algorithm bids more aggressively to find users predicted to meet the new, higher ROAS target.
- ROAS: Your average ROAS might dip slightly from its current very high level towards the new target, but it should still meet or exceed the new target.
- Volume & Profit: Total conversions and total profit should increase if the algorithm successfully finds more valuable users at scale.
- Monitor Closely After Each tROAS Increase:
- Give it 3-7 days to stabilize after each change.
- Watch: Spend Level, Avg. ROAS vs New Target, Total Conversions, Total Conversion Value, Overall Profit (Value - Spend).
- Iterate: If spend increases but is still significantly below budget and ROAS is consistently meeting/exceeding the new target, repeat step 2 (increase tROAS target another 10-20%).
- The "Sweet Spot": You want to find the highest tROAS target where the campaign:
- Consistently spends most (e.g., 80-100%) of your daily budget.
- Consistently meets or slightly exceeds that tROAS target.
- Maximizes your total net profit (Value - Spend).
- Budget Increase Becomes Relevant Later: Only after you've maximized your tROAS target and the campaign is consistently spending its full budget while hitting that high target should you consider increasing the budget. At that point, the algorithm is optimized for efficiency at scale, and adding more budget might allow it to find additional users meeting that strict efficiency target. Even then, monitor closely for ROAS dilution.
In Summary:
- Trust your data: Your campaign is proof that high efficiency at low spend is possible. Google's projection model is flawed for tROAS campaigns aiming for profit maximization.
- Resist the budget increase siren song: It's the wrong tool for scaling efficiently right now. It risks destroying your great ROAS.
- Pull the correct lever: Increase your tROAS target. This aligns the algorithm's goal (find higher-value users) with your goal (spend more profitably).
- Be patient and incremental: Small target increases, monitored over several days, are the key to unlocking scalable, efficient growth.
You've correctly identified the paradox – Google's suggestion doesn't align with your observed performance or your bidding strategy's logic. Stick with the tROAS target increases; that's the path to scaling your success without sacrificing profitability.
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.