We're a long way from the Mad Men days, when advertisers just stood around bouncing ideas off each other before launching campaigns for the masses. Those might be romanticized times, but advances in computing and the sheer volume of data available today mean marketers can use machines to run far more efficient campaigns than ever before.

And yet there's still something incredibly important that online marketers can learn from Don Draper and his ad team. With so much information coming from so many directions--Web traffic, click-through rates, sales transactions, and conversion paths--it's all too easy to put campaigns on autopilot and forget the powerful role that creative people play in the process. The best online marketing campaigns find that sweet spot where technology and human expertise meet, to produce results that go beyond optimization.

Here are four ways that companies should be shaping their campaigns in a data-driven world:

1. Connect campaigns to business goals. Many marketers have a narrow view of lead generation and ROI that focuses on data-driven outcomes that don't help the business grow. For example, a company might want to earn the top spot in organic search rankings for a keyword without knowing how that outcome drives revenue.

Using vanity metrics to define success can lead to campaigns that feel good but don't contribute to the bottom line. Before launching a digital marketing campaign, step back and evaluate the goals of the business. While it may be tempting to choose targets based on what you can achieve using readily available data, not seeing the big picture can be a very costly mistake.

2. Take campaigns to the next level. One of the biggest problems with automation is that it only evaluates existing data and assumes the best advertising copy is already in place, which means it only looks at how well a campaign is performing instead of how well it could be performing. Without human expertise, your campaigns will improve, but they will never grow. The fact that you have so much data at your disposal can cause you to wrongfully believe you have everything you need to be successful.

For example, take a car-insurance company that ranks on the first page of search results for the phrase "car insurance." Though the marketing team might be thrilled, conversions don't follow. As it turns out, most people search for car insurance using geo-specific keywords, such as "California car insurance." A search engine can't figure that out; it takes a team of experts asking smart questions to come up with a better alternative.

A machine can tell you what consumers are doing, but not why they are doing it. In search marketing, the most egregious example of this is in bid management, in which keyword bids are adjusted based solely on their performance. While auto bid management can make existing ads more efficient, it can't use the results to create an even better ad with a different message that will double revenue.

3. Use data insights across channels. An algorithm can optimize ads within one campaign, but it can't tell you how to apply the lessons you learn to other channels. It takes real people to knock down the walls between marketing silos.

Ideally, pay-per-click advertising and SEO should be joined at the hip, but in many cases the two teams never talk. By casting a wide net in paid search and finding a variety of keywords relevant to your business, you can then use that research to set up a better SEO campaign. And as the keywords climb the rankings in Google's organic search results, paid search efforts can be tailored accordingly.

Those lessons in search marketing can then be applied to additional channels, including other display networks, social media ads, and email marketing. It also leads to a robust content-marketing program with sharable stuff created by people for people--which Google generously rewards via better search rankings.

4. Share the data across the whole organization. The lessons learned in digital marketing channels can also be applied offline across the organization. Instead of trapping the data within automated channels, teams should be sharing insights that can help the business grow in other ways.

After the recent data breach at Target, we saw a spike in searches for small business accounting tools that also included security-related phrases. That's the kind of timely information the sales team would love to have while meeting with prospects. Any time new search terms gain popularity, it's an opportunity for a business to use online insights to inform real-world efforts.

At Wpromote, we call the process of combining data-driven decisions and human marketing expertise "intuitive search intelligence." (The concept has been so successful for our clients that we released an e-book that dives deeper into our approach.)

When it comes to marketing, companies don't need to return to the Madison Avenue glory days, but don't forget them as technology takes marketing to places the Mad Men could have never imagined.