An algorithm is only as good as the ground truth it is fed. On day one, your account has no truth—only assumptions. Manual control is how we build the map before we hand over the keys.
There is a common misconception that Google's artificial intelligence can instantly solve all your lead generation problems. I often see new accounts set up with the "Maximize Conversions" bidding strategy from the very first minute. While this sounds like the gold standard, there is a fundamental flaw in this approach for new campaigns: you cannot maximise what you do not yet have.
I recently launched a campaign for a B2B video production studio in the UAE. The market there is bustling, but the competition is fierce. Instead of putting the campaign on autopilot, we decided to fly the plane manually.
We launched using a "Maximize Clicks" strategy combined with strict "Exact Match" keywords.
Why did we do this? ... while Google´s default setting is: Maximise Conversion with Broad match keywords
Because we needed to feed the system valid data before we could ask it to optimise. By using Exact Match for terms like "video production company" or "drone videography," we ensured that every cent of the budget was spent on high-intent searches, filtering out students looking for internships or people looking for movie tickets.
Imagine if we had launched with Broad Match on day one. In the video production industry, context is everything. A broad match for "video production" wouldn't just trigger for businesses looking to hire a professional team; it would open the floodgates to the wrong intent entirely. We would likely be paying for aspiring filmmakers searching for "video production jobs", students looking for "free video editing software", or musicians seeking "cheap music video shoots".
Without historical conversion data to guide it, the algorithm simply cannot distinguish between a marketing director with a corporate budget and a hobbyist looking for a tutorial. Consequently, our budget would have evaporated on irrelevant clicks before a single qualified client even saw the ad.
Expert Insight: The Data-Starvation Paradox
"In the video production industry—and most B2B niches—context is everything. Without historical data, the algorithm cannot distinguish between a marketing director with a budget and a hobbyist looking for a free tutorial. If you launch on autopilot, you aren't flying; you're just paying for the learning phase of a machine that doesn't know you yet." — Dirk Röttges
We also made a conscious decision to disable the Display Network and Search Partners. In a B2B context, these networks often dilute the budget with low-quality traffic from outside the target region. We set the location targeting strictly to "Presence" rather than "Presence or Interest" to ensure we were only paying for clicks from people physically located in the UAE.
The strategy paid off instantly. We saw an 80 per cent impression share on the first day, but more importantly, the business impact was immediate.
The client reached out to me shortly after launch to confirm that the technical reset had worked. They told me, "We are getting some good results, and had 2 new enquiries out of it so far."👍
This is the power of manual control during a launch phase. Because the account is now stable and generating leads, the client is already planning the next steps, looking to potentially run ads in Saudi Arabia. None of this would be possible if we were still stuck in a learning phase with a smart bidding strategy that had no data to learn from.
Once the account accumulates enough conversion data, we will switch to automated bidding. But until then, manual control is the only way to ensure efficiency and immediate results.
If you are launching a new campaign and want to ensure your budget is not wasted on the wrong learning phase, let me help you set the right course.