marți, 3 ianuarie 2012

Tool Review: Wordstream's PPC Grader

Tool Review: Wordstream's PPC Grader


Tool Review: Wordstream's PPC Grader

Posted: 02 Jan 2012 08:48 PM PST

Posted by JoannaLord

In paid search there are a lot of rough Adwords accounts out there. You've seen them I'm sure. I know I've worked on them. We get brought in to assess the situation and make some big decisions around whether an account has potential or whether it would be better just to start fresh. This is one of the most common questions we get asked -- how am I doing?

I'm always looking for tools that can help advertisers answer that question. Usually I have to log in and pull reports myself to get to a point where I feel comfortable enough to jump on a phone call and give my opinion. Needless to say that ends up costing time on both ends which of course costs money on both ends. 

Recently, Larry Kim, founder of Wordstream, dropped by the Moz offices and a few of us Mozzers were able to explore what their team has been working on. We checked out Wordstream's PPC Grader pretty closely and I thought it would be worth running through what we found. I can honestly say it offers us paid marketers a unique snapshot that can be really useful when first getting started with a company or client.

Wordstream's AdWords Tool

Their Google Adwords PPC Grader tool (report w/ some blurring pictured above) runs by looking at your AdWords account and grading against averages and best practices they've seen from analyzing thousands of other AdWords accounts. To help show off some of the tool's benefits I am going to go ahead and share some actual data (and screenshots) of the tool's analysis of our Google AdWords account.

The big deliverable from the tool is a lengthy report page that shows a bunch of data about your account, but the first number you see is a rough "score" comparing your AdWords campaign's performance to other accounts in your spend range. It's worth pointing out that a concern we had with the PPC grader was the "spend range" component. The range SEOmoz fell into was "$15,000 to $99,999,999.00." Our monthly budgets run on the low end of that but we are compared to huge companies with huge branding budgets. You can see where that might affect some data. With that said, the tool is still a great way of showing a client or company the potential PPC has as a channel. It makes the case that with time and resources there is a profitable place to grow this account to. 

Below are the sections included in the PPC Grader (as you can see this is some seriously awesome data!):

  • Account Diagnostics - An overall score and rundown of your account compenents & counts.
  • Wasted Spend - A look at your negative keyword list and projected spend waste calculation. 
  • Quality Score - A graph of your quality score distribution, which is often overlooked by too many paid search marketers.
  • Impression Share - A share of voice view that helps you understand how often your ads are appearing for relevant searches.
  • Click Thru Rate - Analysis to help you compare your CTR to competitors in your spend range for both Search and Display networks.
  • Account Activity - A snapshot to help you better understand the volume of recent activity in your account and comparitive views.
  • Long-Tail Keyword Optimization - Helps you understand if you are holding your own in the long-tail arena, which holds huge potential for PPC-ers.
  • Text-Ad Optimization - Quick glance of how many text ads you have running, and how many per ad group to better understand how tailored your ads are.
  • Landing-Page Optimization - Helps you see how many this account has, and how that compares to your competitors in this spend range.
  • PPC Best Practices - A pass/fail audit whichs helps you quickly identify any low hanging fruit to go after on the best practices list.
  • Overall Grade - A reminder of your account grade and performance summary. 

Phewwww. That sure is a lot huh? This snapshot is an epic first view for paid search marketers. I'm not going to go through each of the sections in depth because that would take forever, but I thought I'd run through some of my favorite pieces to start. After that I'm going to mention some other feedback and feature requests we have. Let's start off with what we think is really awesome...

PRO #1 : Depth & Breadth of Summary

A report with both such a broad view and such a deep view is really hard (if not impossible) to come by in PPC. One of my favorite sections of the Wordstream PPC Grader is the PPC Best Practices checklist. It gives you a top-level view of how well you are doing. Below you can see what came back when we ran the SEOmoz account through the Grader:

PPC Best Practices Analysis

You can see we are doing well on every section, except for "Modified Broad Match Type." This is 100% accurate, we have yet to play around too significantly with match type strategy at SEOmoz, and have instead focused on long-tail expansion. This is a great reminder that we are leaving an excellent option on the table, and we can better prioritize match type expansion in 2012.

PRO #2: Impression Share View

I love this section. This view helps us understand how often our SEOmoz ads are showing up. We can see if our budgets are where they need to be and if our ad ranks are sitting in the right places. This can help with a really common conversation paid marketers have with advertisers around budget. Most  times advertisers want to improve PPC performance without increasing budgets. While optimization should always be happening, there is a time and place for budget changes. Bid management and budget management need to be revisited often. Wordstream's Grader helps make the case for this. 

Impression Share view

You can see that SEOmoz scored in the 58th percentile which is pretty decent, but there is still so much more we can be doing. This sort of view helps us make a case for more budget, as well as helps us prioritize some ad position management internally. This means we need to set aside more time to work on quality scores, and targeting. A view like this can tell us so much.

Okay those are just a couple of the PROs we found by running the Wordstream Grader for the SEOmoz PPC account. I do want to run through some of the concerns we had, and lace in some feature requests the team over here would love to see.

Concern #1: Competitive Spend Range

I mentioned this above but since the competitive bucket we are in is so large, we really need to take this audit with a grain of salt. The budget spend of an account is indictive of big differences in how an account will perform, and be managed. Let's look at an example:
 

CTR Analysis

If we were to just glance at our CTR summary without considering the above limitation we would think we are doing horribly on our CTRs. It shows us performing at the 10th percentile. Yuckkk. If you take a closer look though, we are at a 1.39% CTR and our "competitors" at a 5.34%. The baseline for CTR success is a 1.00%, so we feel pretty great about out 1.39%. Seeing that our competitors have a 5.34% makes it quite clear we are being bucketed with advertisers who have large (in the millions) budgets who likely bid on brand terms (in the hundreds). We only have a few brand terms and those perform at the 4.00% level. 

Feature Request #1:

It is absolutely awesome that Wordstream has collected these audits across thousands of advertisers, but we would love to see a budget drop-down filter that had more bucket ranges to help us better categorize ourself appropriately. As a side request to this it might be cool to see some place where we as advertisers can add in "how long this account has been active" to help further categorize the advertisers so the comparison carries more description. It would be cool to know if that average lifetime of the accounts I'm being compared to is 6 months, or 3 years, you know? Just some ideas. 

Concern #2: Industry Non-Specific Data

Just like in SEO, PPC performance varies greatly depending on both your industry and your site's purpose. By not having an "industry or site type" filter in the Grader, all of my data is compared to advertisers with such different goals, resources, etc. Below is a great example of how this can hurt the data cause:

Landing Page Optimization

Knowing how I am doing on my landing page optimization (LPO) is key. I really appreciate Wordstream building this in, as I believe it's one of the most valuable pieces to PPC success. Unfortunately given the data is not industry specific, and the competitive spend range issues, this data isn't very useful for me.

It shows me with 23 landing pages, and my competitor having 543 landing pages. Having over 20 unique landing pages given our limited PPC spend, and the size of our keyword list is pretty good. That is the score I'd love for them to show me somehow, like a "size of account" to "#of landing pages" ratio. By showing me that my competitor has over 500 pages built out, its safe to assume I am being compared to an ecommerce site, or at the least a site with big budgets and lots of resources (most likely a CMS that has a PPC template system in place). I can't really gauge from this if I am doing well in LPO.

Feature Request #2: 
I'd love the ability to bucket my site into an industry like a "software" or "tech" bucket, and then have them show me how I am doing compared to them specifically. At the very least I'd like to be compared to other SaaS companies, or B2B companies. I'd also love a score that shows me something like -- a "size of account" to "#of landing pages" ratio for my account and then shows me industry standard. I am really impresses that Wordstream has included LPO in their audit, but I'd love to see them push this a bit further.

So What's the Good Word?

While the tool has a few caveats, I think it's safe to say the Wordstream Grader is the best of it's kind out there. I haven't highlighted all the features of the tool here (as it would make this a monster of an unreadably long blog post), and I would encourage anyone with a Google AdWords account to give it a spin, particularly since it's totally free. The report is great for quick snapshorts, and benchmarking your progress. I'd love to hear your experiences with the tool and feedback - I'm a fan of the folks at Wordstream and of the metrics-comparison methodology in general. So feel free to leave your thoughts in the comments below!

Sidenote: Wordstream also has a great keyword research tool and Rand is planning to do a follow-up post at some point, so stay tuned for that! He'll be digging into their keyword research tool to gauge its value for SEO folks. If you've got any thoughts, experiences or suggestions there, that's also much appreciated. 


Do you like this post? Yes No

8 Predictions for SEO in 2012

Posted: 02 Jan 2012 02:13 PM PST

Posted by randfish

It's 2012, and that means we get to revisit our expectations for 2011 and prognosticate for the year ahead. In keeping with tradition, I'm first going to evaluate my predictions from last December before determining if I've got the cred to make some for 2012. Here's the rules:

For each prediction, we'll grade using the following points system:

  • Spot On (+2) - when a prediction hits the nail on the head and the primary criteria are fulfilled
  • Partially Accurate (+1) - predictions that are in the area, but are somewhat different than reality
  • Not Completely Wrong (-1) - those that landed near the truth, but couldn't be called "correct" in any real sense
  • Off the Mark (-2) - guesses which didn't come close

The rule is - if the score is lower than +1, I'm not allowed to make predictions for the coming year. Cross your fingers for me!

Last year, I made 7 predictions:

  1. Someone proves (or a search engine confirms) that clicks/visits influence rankings +2
    Both Google and Bing confirmed in 2011 that they use searcher behavior, including clicks, as ranking signals. This prediction was spot on (though, to be fair, some felt that prior statements had already insinuated this was the case).
  2. Google local/maps adds filtering/sorting -1
    This one was almost completely wrong. I expected something more like what Yelp offers (and I thought Google's move to do this in recipe search was the beginning of something broader). Google has added more suggested searches as seen below, which is the only reason I'm giving myself a "not completely wrong."
    Google Local Suggestions
    _
  3. Social search will rise -1
    This guess was also quite nearly off the mark, but Google's move into social saved it, at least partially. Google+ has added a lot more depth of social elements and signals for the engine, and for anyone logged into their Google/Gmail/Google+ account, the prevalence of social results is quite remarkable.
  4. Rank tracking will be possible through the query string -2
    Sadly, this one was dead wrong. We saw rank tracking in the query string first emerge in 2009 and I was sure that Google would roll this out more broadly, but instead we're still getting only 10-20% of search referral strings with rank data included, and the new (not provided) issue has made manual or machine-based rank tracking even more essential. Sad, because I think this was a big opportunity for Google to be more open.
  5. Mobile will have a negligible effect on search/SEO +1
    While many pundits will surely claim that 2012 will (finally) be the year of mobile, I'd say 2011 has helped prove that the search world  is pretty device agnostic. Rather than changing SEO, mobile and tablet adoption has merely meant that there's more searches around local and location and that the web as a whole is a bigger part of people's lives than ever before.
  6. Software will become an SEO standard +1
    This one's hard to quantify, but I think it's directionally accurate. Here's the Forrester Interactive Marketing report, which notes a large adoption of SEO software at the enterprise level, and with the death of Yahoo! Site Explorer, software and tools from third parties is more essential than ever. I'm not going to give a +2 as I'd say we're still missing conclusive proof that software is "standard," though our upcoming industry survey may help shed light on that.
  7. We'll start to move away from the title "SEO" to something more all-inclusive +1
    It didn't happen in a big way, but the phrase "inbound marketing" and "inbound marketer" appears to be gaining traction. I like the wording, which suggests earning people's trust and interest rather than buying it and includes SEO, social media, content marketing, blogging and web analytics. In our recent survey of agencies, "inbound/organic" agency was how the largest group of respondents described their firms:
    Survey Data
    We'll be releasing the full data tomorrow night on the blog - stay tuned!

When we tally up the numbers, it's +5 and -4, leaving me with +1, just barely enough credibility to make predictions for another year :-)

This year, I'm making 8 predictions (rather than 7). The goal with each is not just to share an opinion, but hopefully to provide some action (implied or explicit) for marketers on at least a few. I'm also aiming to have each prediction be verifiable at year's end, so that I can, once again, check my work.

Prediction #1: Bing Will Have a Slight Increase in US Marketshare, but remain <20% to Google's 80%+

According to Comscore, Bing + Yahoo! have ~30% market share in the US to Google's ~65%. I personally think these numbers are relatively bogus and put much more faith in those generated by sources like Statcounter (which look at traffic sites receive rather than queries performed by a sample audience). Statcounter shows Google at ~82% and Bing+Yahoo! totally to ~16%. I'm guessing those numbers will be pretty similar come January 2013.

Statcounter Google vs. Bing in North America

The biggest reason, IMO, isn't necessarily just brand loyalty and inertia for Google, but their continued superior performance on long tail queries (note: plenty of the comments in the linked-to Reddit thread are worth a read to get a sense of how "early majority" searchers feel).

Prediction #2: SEO Without Social Media Will Become a Relic of the Past

Already, we're seeing SEO and social media marketing become intrinsically intertwined, but in 2012, I believe we'll see SEO without social fade, just as SEO without link building did from 1999-2000. It's not just that social signals are making their way into the ranking algorithms (in both direct and indirect ways), but also that social is becoming the dominant method of both sharing and discovery for web users. The link graph will remain useful for years to come, but the social "sharegraph" is chipping away at its ability to illustrate what's new, interesting, useful, relevant and high quality.

This trend could well be part of what finally weakens the title of SEO (though I think the practice/tactic will remain strong) and forces those of us who've used that name to describe our profession for over a decade to migrate to something broader.

Prediction #3: Google Will Finally Take Stronger, Panda-Style Action Against Manipulative Link Spam

One of the major weaknesses of Google (and Bing, to be fair) is their continued over-reliance on links as an overwhelming ranking signal. Just recently, I took up a friend's offer to point some obviously shady links from sites Google should clearly be discounting at several webpages. We saw dramatic results within 24 hours - #1-5 rankings that have sustained for several weeks (more news on this experiment to come). This shouldn't be the case and Google's webspam and search quality teams know it.

Linkspam Panda

In 2012, I believe Google's search quality folks will roll out algorithmic changes in how they value low quality links that help them regain pride in their work. The embarrassment and quality gap caused by linkspam is egregious and, if left to stand, gives competitors an opening while simultaneously weakening searchers' trust in Google's results. Just as "content farms" took their hits in 2011, I think link spam's up for some blows in 2012.

Prediction #4: Pinterest Will Break into the Mainstream

The last 4 years have seen Facebook, Twitter, LinkedIn, FourSquare and Tumblr all break the 10+ million users mark. In 2012, I give Pinterest good odds for doing the same. Pinterest is also the first major social network where the gender balance heavily favors women (which is, IMO, a great thing).

Pinterest Geek Memery Board
(above, my sad attempt at a Pinterest board)

Because of this breakout, don't be surprised to see lots more posts like these showing marketers how to leverage Pinterest to help share their content and find potential customers.

Prediction #5: Overly Aggressive Search Ads Will Result in Mainstream Backlash Against Google

There are some pretty crazy things going on in the search advertising world right now. To wit:

espresso machine paid search results

On my laptop (which has fairly impressive resolution), I can only see a single organic result, and the paid search markup is incredible. Star ratings, seller reviews, prices and individual items, photos and featured brands are all dominating the page.

0% Interest Credit Cards SERPs

Google's own "comparison ads" in the credit/finance world push organic results down even further, as the Google product still allows for three additional full size ad slots above the organic listings.

Work From Home Sales SERPs

Perhaps the most aggressive of all is Google's new ability to insert a logged-in users email address automatically into PPC ads, as pictured above. These are still rare, but I wouldn't be surprised to see them roll out in greater force.

My prediction is that in 2012, we'll see the start of "paid search blindness" being studied, reported and impacting the engines' bottom lines. Organic results still garner 80%+ of all clicks, but that percent has been dropping as Google gets more aggressive with paid search to continually meet earnings expectations. 

Prediction #6: Keyword (Not Provided) Will Rise to 25%+ of Web Searchers

Despite Google's statements that missing keyword data will stay below double digits, I'm predicting that by December of 2012, we'll be looking at a quarter of all searches coming from logged-in (and thus, keyword-anonymous) searchers. Google's working hard to get adoption of Android, Google+, Google Apps and Gmail, all of which will increase the percent of not provided searchers.

Keyword (not provided) for SEOmoz

While I wish this program would roll back (as there's clearly no real privacy risk or they wouldn't provide the data to paid advertisers), Google's the 800-pound gorilla and the marketing field's counterpoints, while far more valid, likely won't play as well in the media. Google's got the politics sewn up on this one, so our only hope is that they decide to do less evil. Unfortunately, that's not the way they've been trending of late.

Prediction #7: We'll See the Rise of a Serious Certification Program

The search/inbound marketing industry is in sore need of a program that helps early talent in the field become mature professionals. Today, SEMPO, Market Motive, Inbound Marketing University (from Hubspot), Search Engine College and a variety of others provide this service, but none of them are yet at scale or universally respected by hiring managers and companies in the field.

It's hard to quantify what "the rise of..." means. Thus, I'll predict that by year's end, at least one industry certification has 5,000+ users on LinkedIn (currently, Market Motive leads the pack with ~1,700)

Prediction #8: Google Will Make it Very Hard to Do Great SEO Without Using Google+

Google's just started to add Google+ brand pages in search results, They're leveraging Google profiles for rel=author tags. They've made Google+ circles and +1s visible in SERPs. In 2012, I think this pattern becomes a concerted effort by Google to tie promotional efforts in organic results to the Google+ login/verification system. This will not only encourage/force usage of their social network, but give them a much greater ability to tie social, ranking and visibility signals together (and probably fight spam + manipulation, too).

The positive for marketers is that closer integration with the social platform will reward those who can successfully manage both SEO and social media marketing. It's also (hopefully) going to be a boon for white hat tactics that help build brand signals while reducing the effectiveness of exploits that manipulate (like exact match domains or anchor text link spam). The negative is that Google's probably going to get even more data about ALL of our online behaviors, making themselves an even more overreaching and powerful force on the web than they are today. We just have to hope they'll also become more benign, though more power rarely leads to less corruption.


I'm looking forward to hearing your predictions (and opinions on the above) in the comments. I think we've got an exciting year ahead.


Do you like this post? Yes No

Niciun comentariu:

Trimiteți un comentariu