10 Areas for Pay-per-Click Improvement

10 areas for PPC improvement - title slide

TLDR

At any stage of PPC work, early or developed, and at any budget, micro or large, there are things to improve to get the commercial results you are trying for.

But the way PPC optimization is worked on is usually too narrow if there are not specialists with the time and expertise assigned to it. This is the main reason why PPC campaigns "burn cash" without results and often get shut down in frustration. The problem is a lack of exploration of the potential, more often than a lack of potential in the market.

A more comprehensive view will help marketing managers or managers working with agencies define ideas about what CAN be done, and prioritize based on available resources what SHOULD be done first.

The 10 PPC performance areas are as follows:

  1. Campaign targeting
    - advertising to the right people
  2. Keywords
    - advertising at the right time and place
  3. Ad copy and matching
    - being relevant to where you are advertising
  4. Landing page quality
    - keeping people's attention after the click and providing them with value not just spam
  5. Landing page matching
    - over time you should develop more and more specialized landing pages to match different people, keywords, and interests/concerns
  6. Call to action
    - your landing pages have to perform well BECAUSE they provide a compelling proposition that the customer is happy to go along with
  7. Analytics
    - you need to understand what is happening in the connection of the above performance areas both quantitatively and qualitatively
  8. Bid amounts and position
    - you should seek out sweet spots and adapt your bid/cost ideas separately according to different campaigns and past performance
  9. Targeting matching
    - you should improve your understanding of the customer and her concerns in order to feed back into fresh thinking about finding similar people
  10. Prospecting or mining
    - you should make clear decisions, based on most likely return on resources, about whether to dig deeper into existing campaign optimization, or run experiments to discover potential new performance areas

Not just buying any old clicks: The pay-per-click action chain

A successful pay-per-click action chain goes something like this, taking Google AdWords paid search for the example:

  • Someone searches
  • They see your ad
  • They click your ad
  • They land on your landing page
  • They are delighted
  • They take action (and it works)
  • Now or later this action leads to (or contributes to) sales revenue

Doing well in ALL of the above parts of the action chain is far harder than most businesses are able to solve without specialist marketers and significant investment of time, click budget, and resources for creative content development.

Let's break this down a bit more fundamentally:

  1. Person A wants to achieve result B so she searches Google using search term C
  2.  She sees an ad about X which seems to be pretty much exactly what we needed in relation to C - so she clicks it
  3. Arriving on the website (landing page) it is immediately obvious that this page is about the same things in advert X which matches search C
  4. On looking at the page it seems there is something that successfully connects search C with result B
  5. She finds out about product Y which definitely seems like it is going to be good to help with desired result B
  6. Brand Z sounds good and there is an action available from them which helps with result B via product Y
  7. She takes action and (now or later) product Y delivers result B
  8. Person A and Brand Z move to the next stage of the relationship! (The Second Moment of Truth.)

In the above model:
A = a person
B = their concern (problem / goal / task / want / need)
C = the search term they use (keyword /  phrase / query)

Somewhere between C and B you have the person's intent and expectations when performing the search. This is commonly in flux as they adjust the search, get suggestions from Google, refine based on search results, re-search, and develop their line of thought. So a search impression and click may emerge from a fuzzy cloud of intent and trial-and-error rather than a clear "search for C, type in C, click best result for C" -- that is way too simplistic nowadays.

Anyway, it is clear that our ad performance depends on understanding and matching what the customer wants. The degree of relevance / match is the main factor in getting clicks and holding initial attention.

In the above model we had:
X = your advert and landing page combination
Y = your product or service
Z = your brand

From your (business) point of view your instinct is to frame your advertising as "We are Z and we offer Y, bamm here is our advert X! Buy now!"

As we can see from the customer's experience, what we should be doing instead is: "Advert X because you are looking for search term C! Here is how product Y is great for when you want to solve B! By the way brand Z really reflects your values, Ms A."

In pay-per-click advertising you therefore proceed in the following order of improvement:

  • Match your ad X to search term C
  • Match your product/service Y to concern B
  • Match your brand Z to person A

As an advertiser, your long-term goal is to improve your understanding of the customer, but you have to get there via improving the message of how great your product/service is versus the customer's wants and needs, and before THAT you have to get them into the virtual sales message via understanding how to match ads to search terms via interpreting users' intent.

If your understanding is weak about person A's desire/need for B as hinted at via their search term C, then you are not going to be routinely successful in selling to them, let alone creating recognition or trust in your brand.

This is pay-per-click advertising. It is NOT just keywords, ads, bid amounts, and cost/click statistics. It is the whole picture of how to match a customer who might like you, with a product you can persuade them will be great (and is), and catch them at the right moment to gain and hold attention to pitch that product. It is an encapsulation of the whole of your marketing and business proposition, in miniature, in a controllable sequence.

Any part of the sequence that is off, will diminish your results.

The better each part works, and the better it all matches and works together, the better your results. Therefore PPC improvement has to keep touching all parts of the action chain for your customer's experience and how your advertising serves that.

It is important to take this deeper view of what PPC advertising is all about because a shallow focus means that your optimization ideas are too limited to make a difference. It is very common for PPC advertisers to be stuck trying to acquire clicks and then try to do better at getting clicks, and then feel frustrated with a lack of commercial results. It is too easy and tempting to do a lot of quantity (or even accuracy) of matching your ad X to search term C... but completely failing to work on matching your product Y to that person's concern B. And because you are self-centred on your brand Z you are not even clear who the person A is that you ought to be speaking to in your ads and offers.

It is on this above strategic basis KnowledgePower works on the basis of 10 improvement areas in PPC.


10 PPC Improvement Areas

slide from video presentation showing the same 10 ppc improvement headings

  1. Campaign targeting
  2. Keywords
  3. Ad copy and matching
  4. Landing page quality
  5. Landing page matching
  6. Call to action
  7. Analytics
  8. Bid amounts and position
  9. Targeting matching
  10. Prospecting or mining

The video presentation outlines what is meant by each of the 10 areas.

Here are the headings and some example problems to illustrate further.

1 - Campaign targeting
Locations, radius, exclusions
Bid adjustments
Budget allocation
Schedule
Demographics, audiences
Placements (general)

Example problems:

  • Too wide locations, not relevant enough to the business
  • Targeting relevant locations but sharing too-small click budget between locations where it could be better spent on a more focused area
  • Bidding the same amount for nearby and far away clicks even for a business which is more likely to convert local customers
  • Advertising in irrelevant display placements, especially apps

2 - Keywords
range
specificity
negatives
intent
analysis
research

Example problems:

  • too many auto-suggested keywords
  • not enough lateral-thinking keywords
  • over-reliance on broad search terms in the long term
  • lack of clarity about meaning of search terms (intent)
  • no sources of keyword ideas apart from Google tool

3 - Ad copy and matching
Are your ads, ads?
Split testing
Extensions
Intent matching
Motivation matching
Differentiation

Example problems:

  • ads are boring
  • ads are not specific to the keyword
  • extensions not used
  • ads don't provide strong incentive to click
  • ads look the same as all the competition

4 - Landing page quality
Millisecond matching
Deliver value
Usability
Skim / Sort / Optional depth
What’s next?
Background signals

Example problems:

  • landing page isn't that connected to what the ad said
  • landing page doesn't appear connected to the person's original search and expectation
  • landing page is hard to read or use
  • even if generally relevant, hard for user to move to more specific information
  • no clear next step after skimming or reading
  • poor visuals or lack of visuals
  • poor style or lack of positive brand image

5 - Landing page matching
Closer intent matching
Closer concern matching
Unique media
Brand
Tailored CTA
Conversation / Trial

Example problems:

  • landing page only offers superficial customization for audience
  • lack of a sense that you understand customer's concerns
  • lack of detailed or sufficiently targeted information or utility for the niche audience
  • overly generic call to action or offer
  • lack of a logical next step in connecting with product / brand

6 - Call to action
Why
Why now
Reduce friction
Smooth the way
Provide brand-laden value
Test but preferably avoid guessing

Example problems:

  • lack of compelling enough offer
  • lack of urgency
  • complicated or confusing action process
  • failure to reduce uncertainty / nervousness about contact or action
  • call to action is just a marketer's guess and fails to take account of real users' concerns

7 - Analytics
Know what you are trying to discern
Keep things easy
Track micro conversions
Track the other side of the conversion
Don’t ignore qualitative and anecdotal research
Don’t ignore the possibility of asking humans

Example problems:

  • focusing on metrics rather than questions the metrics can answer
  • process of collecting and interpreting data is a waste of time compared to the value
  • analytics provide interesting things but not cause for action
  • tests are not actually tracked
  • over-reliance on gross signals bounce rate and final conversion without middle indicators being tracked
  • reliance on only website analytics instead of potentially more revealing insights from reviews, orders, complaints, sales staff, customer interviews, etc etc

8 - Bid amounts and position
Have a target maximum cost
Fine tune
With sufficient data, test position
Watch quality score
Watch display position
Different bid strategy depending on confidence

Example problems:

  • maximum cost per click is not set according to any stated criteria
  • inflexibility about testing higher cost clicks or about targeting high positions
  • using the same bid on totally different kinds of campaign
  • lack of quantity of data to support ad position / cost balancing
  • not sticking with campaigns (and working on them) long enough to attain sufficient confidence to optimize bidding

9 - Targeting matching
Budget allocation by confidence
Seek out catch-alls
Seek to understand intent and concerns
Audiences
Move further up and down the funnel
Split test campaigns by targeting

Example problems:

  • division of budget by neatness instead of marketing/sales priority
  • too small overall budget split between several campaigns
  • no link between budget allocation and targeting of campaigns (e.g. prioritize local)
  • no use of remarketing
  • campaign targeting decisions unsupported by testing

 

10 - Prospecting or mining
Discovering new themes
Digging deeper into good ones
Trying to change weak ones
Different media
Different networks
Different incentives

Example problems:

  • not having any other network ad campaigns or tests so no ability to compare AdWords costs against alternatives
  • text-only adverts
  • lack of development of graphic display ads
  • hopping to new campaigns before existing ones are fully explored
  • conversely, doggedly trying to fix failing campaigns instead of looking for easier wins somewhere else
  • working a lot on ad tests and new campaigns but not being creative with conversion offer

The 10 areas as an agenda

You can ask the following 10 PPC improvement questions to yourself FOR EACH CAMPAIGN you are running, or use them as an agenda for discussion with your marketers.

For each question, consider whether there are obvious things to fix/improve, and whether there are ideas you can test... then take your ideas and prioritize into a to-do list for the next month or two. In this way every campaign gets optimization work across its whole performance spectrum, whatever its current stage of development, and it gets this optimization continuously because none of these areas will ever be perfected. It is just about which improvement opportunities offer a potentially good return on investment.

  1. What are the current campaign targeting settings exactly, and why?
  2. Are we targeting all of the best keywords?
  3. Are our ads doing well for our selected keywords?
  4. Are our landing pages holding attention well?
  5. Are our landing pages providing unique value to different customers with different concerns?
  6. Are people taking our desired action on/from the page? Is our desired action a good match for what they want to do?
  7. Are our analytics providing us easily with insights about where to improve and what test changes are helping?
  8. Are we getting a good balance of click cost and click volume?
  9. What are we doing to target the best kind of customers more effectively?
  10. Should we be going deeper into these themes now or should we be finding new things to try out?

ALL PPC campaigns should have ALL of these questions considered EVERY time you review and plan improvement work.

Finally, in a smaller organization it is not rational to spend large amounts of time discussing and documenting things on this kind of strategic level. You just want to iterate quickly, control costs, and seek easy wins, using intuition and available expertise to short-cut to the likely effective tactics.

That is fine.

That is where a PPC expert helps you because he or she automatically cycles through these kind of strategic questions - and resource/effectiveness balancing - the whole time in their head, and will help you short-cut to what is most likely to work... if not instantly then with less trial and error, and more potential for future optimization, than you will achieve by yourself.

Read more about KnowledgePower PPC management services.