I’d like to share this report from Marketing Sherpa titled ‘Special Report: Online Keyword Research Guide: 5 Tips & 9 Useful Tools’. I broke it up into 2 parts with the 5 tips coming first.
Takeaway #1. Tools can’t replace your own research
You can’t build a search campaign solely on the terms generated by a keyword research tool. Your competitors are probably using the same tools as you are using, and can find the same terms.
Instead, you need to use those tools to augment the proprietary keyword research you’re performing for your company or clients:
- Look carefully at your own product names, popular terms used in your website copy and page titles, and internal site search queries terms.
- Study existing marketing materials, press releases and other content for important industry terms.
- Conduct your own online research on user forums, blogs and social media sites for terms and topics that users are talking about.
- Perform your own competitive analysis of how other companies are positioning their sites and search ads.
After completing your research, use a keyword research tool for the following tasks in generating your overall list:
- Starting a keyword list for a new product or new ad campaign.
- Seeing additional, related terms to supplement your internally generated lists when you’ve exhausted your own creativity.
- Checking your existing list for estimates of search volume and competition for PPC and organic placement.
- Gaining insight into unusual search syntax based on historical queries, which can help you generate new search phrases.
Takeaway #2. Data isn’t 100% accurate
The reliability of data is the most common caveat we heard from marketers who use online keyword-research tools.
The data these tools generate is typically only a snapshot of overall search activity, which depends on:
Source of the data (one search engine vs. several; ISP data vs. proprietary Web scraping tools, etc.)
Frequency of update of data
“Volume estimates should always be taken with a grain of salt,” says Greg Hayes, Online Marketing Specialist, G.A.P Adventures. “They could be low or high depending on the day they were collected.”
Data accuracy can be a bigger issue as your search criteria become more granular. For example:
- Niche terms and long-tail keywords that have lower search volume may be less well represented in keyword databases. That was an issue for Bill Shaughnessy, Director, Online Marketing, EDGAR Online, as he conducted keyword research for long-tail financial search queries.
“The challenge I faced was with some of those less popular, less searched terms that an equity analyst might use, such as ‘derivative’ or ‘derivative models,’” says Shaughnessy. “I thought that [the tools] were lacking, because they kind of led you back to broader-based terms that I already had.”
- City-specific or regional data for geotargeting might not be well represented in keyword databases due to the lower volume of that search activity captured by the tools.
“There’s not enough data to make it statistically relevant,” says Melissa Mackey, Online Marketing Manager, Fluency Media.
Takeaway #3. Search advertising cost estimates are directional at best
Several tools will give you estimates for the average cost per click companies are paying on top keywords. But that data can’t tell you exactly how much you can expect to spend on a particular ad campaign.
For starters, many cost-estimating features only provide a cost index, showing whether the term in question typically commands a high, medium or low cost per click. Tools that provide actual price estimates typically display results in a range — for example, $0.50-$4.50 — that can lead to a wide variation in your potential budgeting needs.
Don’t look to tools for hard and fast CPC spending guidelines. Use them to create estimated costs for new keyword lists or develop an hypothesis on how keyword changes might impact your budget.
Takeaway #4. Use multiple tools for best results
Each tool uses different databases and analytics. So, their results will differ. But you can use those differences to your advantage by compiling data from several tools.
“I don’t think there is any one source that can do the job for you,” says Stuart Meyler, Principal, Stinson Partners. “We look at three or four of them to come up with a baseline.”
Paul Elliot, Partner, Acquisition Marketing, Brulant Inc., had his team create special algorithms that reconcile data from multiple research tools.
The more you experiment with different tools, the more you’ll find features and data sets that are most relevant to your business. You may trust one tool’s volume scores, but find better-related keyword suggestions elsewhere. Another might be particularly good at researching competitors’ search marketing activities. Use them all to enhance your strategy.
Takeaway #5. Always test the keywords in your own campaigns
Even the most accurate aggregate data is no match for live test results. Whenever you use an external tool to create a new keyword list, conduct a test of those terms as part of your research strategy.
“We always test a keyword list for 30 days to see if we have a realistic baseline or benchmark for where we should be moving forward,” says Kent Lewis, President, Anvil Media.