The majority of online ads are simply terrible. Marketers generally rely on a system best described as “spray and pray” hoping that more volume can somehow outweigh mediocre and often irrelevant content.

DataPop, on the other hand, has been using advanced data science for more than two years to optimize the online ad content served by its clients. The results, in terms of increase in conversion rates and return on investment, have been dramatic.

The company announced today that it has raised a $7 million Series B round led by MK Capital, with additional participation from each of its Series A investors including Rincon Venture Partners, IA Ventures, Momentum Ventures, and Accelerator Ventures.

DataPop was founded by a team of former Overture and Yahoo ad specialists and uses machine learning and semantic technology to finely tune ads for marketers. Its system combines consumer search intent signals with product data like SKUs, inventory, price, and geographic information. It then quickly tests thousands of subtle varieties of creative content and ultimately arrives at custom ads optimized specifically for each individual online consumer demographic.

For example, middle-aged women in the southwest respond to far different cues than co-eds in Manhattan, even when searching for the same products. DataPop can deliver unique ad content to each audience segment in real time.

Bag Borrow or Steal, one of DataPop’s clients, has seen some pretty awesome results since deploying the optimized ads six months ago. Its engagement statistics doubled and it achieved a 31 percent increase in return on ad spending. At the same time, the company was genuinely surprised at many of the insights revealed which seemed to contradict traditional thinking. For example, ad copy describing the variety of popular color choices available in a line of designer handbags far outperformed copy which highlighted related pricing promotions.

In addition to advertisers, DataPop works closely with a number of digital agencies. I got a chance to speak with one Santa Monica-based RPA, which is a big believer in optimized ads. The agency said that while too many other search management companies have been competing in a race to the bottom on price, DataPop is one of the few innovating on features and performance.

As a result, RPA’s clients have achieved their own impressive results, seeing an average 25 percent increase in ad-ROI post optimization. One less touted benefit uncovered along the way is that, over the long term, this translates to a higher Google quality score, increasing CPA’s and allowing higher ad positioning at lower CPC’s. In non-industry speak, this means RPA’s clients get the same bang for less buck.

One concern I had when talking to DataPop co-founder and CEO Jason Lehmbeck was that the company’s clients would gather the insights provided through initial optimization and then no longer feel the need to use its services. Lembeck, who was less worried, says, “You can take anecdotal learnings from what worked previously, but last year or even last month are a different world than today. Things change at incredible speed and velocity on the Internet, including what consumers are looking for and what succeeds in catching their attention.”

DataPop plans to use the proceeds of its latest investment round to significantly scale its engineering team and to build out its sales force. As successful as it has been to date, its paid search technology only scratches the surface of the total online advertising opportunity.

Anywhere that the company can gather an intent signal, it can use its optimization technology. DataPop is looking ahead at opportunities to deliver creative optimization for display advertising, expand into social networks such as the Facebook Marketplace and possibly even participate in demand-side platforms.

As more and more consumer activity shifts online, there is a global race to effectively capture and monetize their attention. As Lembeck said to me at the end of our meeting, “The promise of Internet advertising is thus far unmet.”

With a substantial technological head start and a war chest of growth capital, DataPop is out to see that this doesn’t remain the case for much longer.