Hannah Smith

Digital Marketing Portfolio

Udacity Digital Marketing Nanodegree - July 2018

Display Advertising

Campaign Evaluation

​

Summary

​

For this project, I was tasked with analyzing an ad campaign run on the Display Network to identify successful elements, evaluate overall performance, and make recommendations to optimize future iterations.

This campaign consisted of both Image and Video ad formats, so I have included a detailed analysis of each.

Display Image Campaign: Overall Results�

​

Find below the overall results of the Display Image Campaign.

Results: ROI

​

Overall, this campaign was a success, with an positive return on investment. The ads should have ultimately generated (estimated based on assumed conversion rate) 4 new sign-ups for the DMND program at a cost of about $218.13 for each new student. This means that each new sign-up brought in a profit of about $80.87, for a total ROI of $323.48.

Display Image Campaign: Ad Results�

​

Review the Ad results below of Creative A and B for Display Image Campaign.

A

​

B

Which Ad Group Performed Better?

​

I believe that the reason that Creative - A outperformed Creative - B was because the images used in A included people and were more colorful, whereas the images in B were of inanimate objects - tech devices - and were less colorful and catching. People are typically more drawn to images and advertisements that include people, especially because they feel familiar and inviting, and present the product in a way that the potential consumer may be more apt to imagine themselves using it as well.

�

Display Image Campaign: Keywords�

​

Review the Keyword results for the Display Image Campaign

​

Full keyword list in separate file here

​

Image Campaign: Keyword Performance�

Suggestions for Campaign Optimization

Suggestion 1: Technically, when analyzed individually, none of these keywords were all that successful against the overall marketing objective of generating new signups. When rounding the number of new signups generated, every one resulted in 0 new signups. The overall campaign metrics should be taken into consideration when evaluating the performance of individual keywords, as there are some discrepancies due to rounding. In cases where there were 0 (less than 1) new sign-ups generated, I can also consider number of clicks as a metric of success. Therefore, I’d recommend removing or pause the following 3 keywords, as they had a negative ROI and a poor CTR. I would generally recommend optimizing the keyword list for this campaign, however, as outlined on the next slide.

-marketing careers�-marketing program�-digital media online course

See here for my full keyword analysis and ROI calculations.

​

��

Suggestions for Campaign Optimization (cont.)

Suggestion 2: Use AdWords Keyword Planner to research and add new keywords similar to the current top three performing keywords:

1) marketing online�2) online marketing course

3) marketing courses

​

You can view my initial forecast analysis for my keyword research here.

​

From this initial list and after analyzing, I could only recommend two keywords that would yield a positive ROI (“google marketing course” & “net marketing”). However, the extent of their positive return relative to the existing keywords is unimpressive at best, so I might not recommend that either be added to the working keyword list. Further iterations of this process of research and analysis would likely yield some fruitful keywords, however, so I can recommend this process at least.

Suggestions for Campaign Optimization (cont.)

Suggestion 3: A/B test a new description with the more successful creative (Ad A).

​

Current:

Become a Digital Marketer

Become a Digital Marketer, learn through real-live advertising projects

Learn from the biggest names in the industry, and gain experience with real-live projects.

Suggested:

Become a Digital Marketer

Learn through experience with courses taught by industry professionals.

Work on live advertising campaigns and build your portfolio as you work towards your new career.

Display Video Campaign: Overall Results�

​

Review below the overall results of the Display Video Campaign�

Results: ROI

​

This campaign was a success, with an overall positive return on investment. The ads should have ultimately generated (estimated based on assumed conversion rate) 4 new sign-ups for the DMND program at a cost of about $206.40 for each new student. This means that each new sign-up brought in a profit of about $92.60, for a total ROI of $370.40.

�

Display Video Campaign: Ad Results�

​

Review below the Ad Results of the Display Video Campaign�

B

A

​

Campaign A - Short Keyword List

Campaign B - Large Keyword List

Which Ad Group Performed Better?

​

Campaign B outperformed Campaign A by a wide margin, with a 4113.65% higher ROI. I believe this is because not only was the keyword list used in targeting this campaign larger and more finely tuned to the intended audience, but it also included many more long tail keywords. For these more specific long tail keywords, the ad was likely a closer match to what information users were after, and as a result clickthrough rates were higher.

�

Display Video Campaign: Keywords�

​

Review the Keyword results for the Display Video Campaign.

​

This is an excerpt of the keyword list, the full list is here

​

​

Key Campaign Results (Keywords)

​

​

I would rank the top 5 performing keywords in the following order, starting with the top performer. This is based first on the number of new DMND signups they were able to generate, and then by their ROI in dollars. When analyzing individual keywords, rounding the estimated number of signups to the nearest whole number results in indicating that all but the top 2 keywords generated 0 new signups. It also results in showing the total number of new signups generated as 2, instead of 4 as calculated previously. The overall campaign metrics should be taken into consideration when evaluating the performance of individual keywords, as there are some discrepancies due to rounding. In situations where there were 0 new signups generated for an individual keyword, I measured success solely in terms of ROI - those that had the least negative ROI ranking more highly.

​

1. ad advertisement
2. online marketing video
3. avinash kaushik
4. learn digital
5. google website optimization

�

Key Campaign Results (Keywords, cont.)

​

The Keyword “ad advertisement” was the most successful keyword in this campaign. Because the marketing objective for this campaign was to drive signups for the DMND nanodegree program, I would consider only those keywords that generated, based on our assumed .2% conversion rate, at minimum 1 or around 1 (when rounded) to have significant success. Based on that reasoning, the top two performing keywords would be “ad advertisement” and “online marketing video”. Of those two, “ad advertisement” had a significantly higher ROI of $210.88, compared to $43.65 for “online marketing video”.

See my full analysis and calculations here.

Keywords Performance Snapshot

Suggestions for Campaign Optimization

​

Suggestion 1: Based on the marketing objective of generating new signups, there were not many keywords that were wildly successful. I would perform keyword research in Keyword Planner to find and add new keywords that were similar to the top 5 performers previously mentioned and add those to the list for the future.�

​

​

��

Suggestions for Campaign Optimization (cont.)

​

Suggestion 2: I would also remove or pause any keywords that had a negative ROI, such as:

�

social media training�internet marketing�google advertising�google analytics�digital marketing social media�digital marketing

​

business training course�Digital Marketing�social media marketing�search engine marketing�advertisement google�google seo

See my full analysis and ROI calculations here.

Suggestions for Campaign Optimization (cont.)

​

Suggestion 3: Since initially for this video campaign the creative was kept the same and used for two different keyword lists, I would implement the previously described keyword optimization tactics and then A/B test creatives for this new and improved keyword list. Variations that could be tested might include alternate video lengths, such as longer or shorter versions of the same creative - :15, :30, :60, etc. I could then assess the view rate for each along with the other important engagement metrics to determine which version performed best. This would probably be the best place to start, rather than creating an entirely new video, in order to save time, resources, and most especially to keep costs down.

​

​

Recommendations for future campaigns

​

It does not seem to me that this campaign was set up with any customer segmentation to distinguish creative and ad formats. Depending on how much additional budget is allotted to use in continuing this campaign, I might also build out different target personas and tailor ad groups and their component ads, keywords and creative towards each specifically.

Since I do not know how much additional budget this will be given, I would instead recommend beginning by optimizing the existing ad groups, ads, creative and keyword lists for the best performance.

​

Ads

I would begin by selecting the most successful of the image ads, in this case, Creative A, and run only that moving forward for the image campaign. Since I have enough data to determine that this ad was more successful and had a better ROI than Creative B, I would pause Creative B and focus my funds and attention on making A that much more successful.

This could be accomplished by optimizing the keyword list for this ad. Since the video ads featured the same creative, and differed only by which keywords were used, I can take the data from that campaign into consideration when building an optimized keyword list.

​

Keywords

Now that I have some data on keyword performance, I could take the best of each keyword list from both the display and video campaigns to build an optimized keyword list to use across all of the ads.

While each of the three lists had some high performing keywords, there were also plenty that did not perform all that well, as well as others still that had a negative ROI. I would definitely recommend removing or pausing any keywords with a negative ROI, and since these keyword lists do not seem to be set up with any specific customer segmentation in mind, I would also recommend compiling the best from each list into a single working keyword list to use for both the image and video ad formats.

Based on my analyses and calculations outlined in the links provided on previous slides, I would pause all of the keywords on this list.

​

​

Working Keyword List - Revised

​

ad advertisement

online marketing video

digital learning courses

course digital marketing

seo

digital business course

Learn facebook advertising

Adwords course

online marketing

seo marketing

digital marketing course

digital internet marketing

SEO training

marketing certificate programs

Digital Marketing Training

digital online

online marketing training

how to optimize a website for search engines

google advertising services

google analytics training

​

​

​

display advertising course

google search seo

Social Media marketing training

avinash kaushik

learn digital

google website optimization

Mailchimp training

business marketing courses

google analytics course

courses in social media

internet marketing seo

advertising marketing

google ppc advertising

marketing online

online marketing course

digital marketing training

marketing courses

Keywords (cont.)

Depending on the additional budget allotted for continuing this campaign, I might also be able to put more money behind the bids for the top performing keywords. For now, I would suggest leaving the bids as they are, and standard across all keywords in this new list to see how they perform.

A/B Test

I’d recommend implementing an A/B test for the video creative to see if a different length of creative might perform better.

​

In the original campaign run, the video creative was :41, and the same version was used across two different keyword lists. Now that I have optimized a keyword list as mentioned in the previous slides, I would test a shorter video to see if that might improve the view rate and ultimately, the conversion rate.

​

I would do this by splitting the campaign into two ad groups - one for the original :41 creative, and a second for a :30 creative. A difference of :11 may not sound like much, but I believe that with such high competition already for the viewers’ attention, and with the temptation to “skip” or click away from advertisements in general being so strong, the sooner I can get the ad’s message across, the better.

​

I would keep the keyword lists the same for both creatives, so I could then analyze results and determine at the end of the next run, all other things being equal, which creative was more successful.

​