AdWords and Excel are an inseparable combination for any SEM professional. PPC Hero's Eric Couch dives in to advanced Excel tactics that can turbo-charge your account performance, including Conditional Formatting, VLOOKUP, and Excel Solver.
Report
Share
Report
Share
1 of 81
More Related Content
A Match Made in Heaven: AdWords and Excel - Hero Conf 2014
1. A MATCH MADE IN HEAVEN:
ADWORDS AND EXCEL
APRIL 28, 2014
April 28-30, 2014
2. WHO ARE YOU?
ERIC COUCH
SENIOR PPC ACCOUNT MANAGER,
HEAD OF TRAINING
@HANAPIN MARKETING
WRITER @ PPC HERO
www.ppchero.com
@ecouch11
3. ME, MYSELF & EXCEL
ME
EXCEL
(ACCOUNT WORK)
@ecouch11
6. THE ANSWER
AN EXCEL FUNCTION THAT
LETS YOU FORMAT THINGS.
CONDITIONALLY.
@ecouch11
7. THE (REAL) ANSWER
CONDITIONAL FORMATTING IS
A FUNCTION THAT ALLOWS
YOU TO AUTOMATICALLY
CHANGE THE FORMAT OF A
CELL BASED ON THE VALUES
AND PARAMETERS THAT YOU
DICTATE.
@ecouch11
8. TEXT COLOR AND FORMAT.
CELL COLORS
BAR GRAPH OVERLAYS.
INSERTING GRAPHICS!
WHAT CAN IT CHANGE?
@ecouch11
20. WORKING SMARTER
YOU CAN DO THAT, BUT WHY
WOULD YOU WANT TO?
THERE’S A BETTER WAY.
AND THAT WAY IS
CONDITIONAL FORMATTING.
@ecouch11
21. A NEW PROBLEM
LET’S SAY YOU NEED TO
GENERATE 349,372 UNIQUE
KEYWORD-LEVEL
DESTINATION URLs.
THEY HAVE UNIQUE
LOCATION, AD GROUP, AND
PRODUCT IDENTIFIERS.
@ecouch11
24. WHAT’S THAT?
VLOOKUP = VERTICAL LOOKUP
BASICALLY, IF YOU NEED TO
CROSS-REFERENCE A TON OF
DATA…
THIS IS HOW YOU DO IT.
@ecouch11
25. WHAT’S THAT?
CONCATENATE = COMBINES CELLS
IF YOU NEED TO COMBINE
MULTIPLE CELLS…
HINT: CELLS THAT YOU MIGHT
HAVE POPULATED WITH VLOOKUP.
THIS IS HOW YOU DO IT.
@ecouch11
43. WHAT ABOUT THEM?
IS THERE A WAY TO
INTELLIGENTLY DETERMINE
HOW TO BEST SPEND OUR
ADVERTSING BUDGET?
CAN WE DO THIS ACROSS
MULTIPLE ACCOUNTS?
@ecouch11
49. WHAT’S THAT?
EQUATIONS LIKE “WHAT’S MY
MOST EFFICIENT BUDGET
ALLOCATION TO MAXIMIZE
CONVERSION VOLUME”.
IT’S NOT INCLUDED BY
DEFAULT, BUT YOU CAN FIND IT BY
GOING TO YOUR EXCEL ADD-ONS
MENU.
@ecouch11
52. EXCEL SOLVED BUDGETS
@ecouch11
A MIX OF BING
AND GOOGLE
CAMPAIGNS
AVERAGE DAILY
SPEND ACROSS
ALL CAMPAIGNS
OPTIMAL BUDGET
AS FOUND BY
EXCEL SOLVER
SOURCE: SAM OWEN
PPC HERO – “HOW TO USE EXCEL SOLVER TO OPTIMIZE YOUR CAMPAIGN BUDGETS”
http://www.ppchero.com/how-to-use-excel-solver-to-optimize-your-campaign-budgets/
65. HOW IT WORKS
SOLVER WILL TAKE OUR
BUDGETS, AND USING OUR
AVERAGE CPC, WILL FIND THE
NEW CLICK TOTAL FOR EACH
CAMPAIGN.
THAT’S SOLVED CLICKS.
(SOLVED BUDGET * AVERAGE CPC)
@ecouch11
66. HOW IT WORKS
USING SOLVED CLICKS, YOU THEN
USE YOUR CONVERSION RATE TO
FIND THE NEW CONVERSION
TOTAL FOR EACH CAMPAIGN.
THAT’S SOLVED CONVERSIONS.
(SOLVED CLICKS * CONV. RATE)
@ecouch11
69. THE TAKEAWAY
JUST BY REALLOCATING OUR
SPEND, WE CAN GAIN AN AVERAGE
OF 1.2 CONVERSIONS PER DAY.
THAT PROJECTS TO 36 MORE
CONVERSIONS A MONTH, A 17.6%
INCREASE WITH NO OTHER
CHANGE EXCEPT BUDGETS.
@ecouch11
Right here– it’s a standard feature in the “Home” ribbon on Excel, and features both pre-made and custom built rules that allow you to format your cells based on the data found within.
These are all pre-built options that you have available to you. If you want to create a new rule, or modify these formatting options you can do so at any time. I won’t go in to the specifics of custom rules, as that’s a presentation all unto itself. I *will* go in to the specifics of how you can create an insightful, actionable report in just a few minutes through the use of conditional formatting. We’ll use the background report as an example – we’ll turn this ugly spreadsheet in to a gorgeous heat map.
This is the same report you’ve seen in the background – just with a few added columns for Average CPC, Conversion Rate and Click-Through Rate. I don’t know about you, but my vision is getting kind of blurry just *trying* to decipher what all that means with that stark white background. You can can certainly find insights here, but there’s an easier way to do it.So to start, you’ll want to highlight the cells you’d like to format. A word of warning, though: Excel won’t make a distinction between columns and their relative scale if you just highlight everything and format it all at once – so formatting both clicks and impressions together will skew your heat map. Nor can it determine if being lower or higher is better for certain metrics – a low CPA is fantastic, but a low Conversion Rate means you’ve got trouble on your hands. Making that mistake will cause your heat map to look like this:
…which is a real hot mess of a heat map. That particular pun was requested by Sam Owen, so I take no responsibility for its’ inclusion. But really, you can’t tell anything from this incorrectly formatted report. The highest numbers, which are Impressions here, are listed as the best. This report doesn’t make any sense! Instead, you should highlight cells 2 through 25 in each column independently and format them based on your success metrics – a higher CTR is better, a lower CPA is better, that kind of thing. Doing this changes our heat map like so:
Now we start to see an interesting picture start to emerge. I’ve left the Clicks, Impressions, Cost, and Converted Clicks columns untouched right now, as I prefer to display those with a different formatting. The red vs. green distinction is a little too arbitrary for these numbers, so I go with a bar graph to show them instead.
*This* chart tells us a story. It portrays the ebb and flow of our daily traffic, and it clearly shows us both where we can pull back on our bids with dayparting modifiers – early morning from midnight to 4 AM – and where to be more aggressive. We have a real lost opportunity here starting at 8 AM – 10 AM to increase our bid modifier and gather even more traffic. Our Impression Share and Average Position ,and Average CPC metrics let us know exactly when our competitors are ramping up their spend, and the hours in which we should do the same… until our Conversion Rate drops off at the lunch hour.This report takes less than five minutes to pull, and no custom modifications were made to these rules aside from highlighting each column individually and selecting the right formatting. Dayparting is the easy and obvious use for this kind of analysis, but it can also assist in ad reviews, geographic performance… really, in any case that leaves you staring at a spreadsheet for hours on end, mining for insights.
This is the same principle applied to a spreadsheet of CPA by hour of day and day of week. I’ve highlighted the areas in which we see specific CPA trends – trends that inform us of which days and hours are places to pull back. Doing this without a heat map would require you to sort through 216 cells of data to make sense of it all.If you’re like me, and were here at Hero Conf last year – you saw John Gagnon’s heat map skills. Now you can do them as well.
“Now Eric,” you might be saying. “That’s an oddly specific number.” I say to you, “you’re quite right.”
We’ve used VLOOKUP for all kinds of tasks, ranging from the following URL example, all the way to cross-referencing average and maximum CPC costs in overlapping/competing accounts. Once you know how to use it, it becomes incredibly handy.
So this is what you see when you type in =VLOOKUP in Excel. It’s kind of intimidating! But it’s also really easy to understand once you’ve tried it a few times. So here’s how it works:
I realize “thing” is a very technical term, but try and stay with me. So for our URL example, the Lookup Value could be an ad group, or a campaign.
Going along with the URL example, the Table Array could be a what you’ve selected out of a master list of Ad Groups and their corresponding UTM tags, or product category ID’s… that kind of thing.
So if, in your highlighted table, your ad groups are in column one, and your UTM tags in column two, you’d list this value as a “2”. Completing this formula is basically saying “When you find this value in this other spreadsheet, and it has this data right beside it, put that other data next to this value here too.”And then I usually ignore the Range Lookup.
Very creative naming conventions, I know.
Doing this kind of leg work beforehand, having a master list like this for each unique element, will make the URL generation process lightning-quick going forward. I’ve created URLs with this method that look up
You can do this for every single unique element of your ad – if you have tracking requirements that also need location information, keyword-level product ID’s, they can all be cross-referenced here and will make a painful process much easier. Otherwise, you’d be stuck filtering and auto-filling URL elements… or even typing manually.
Here, we’re using the substitute function to remove all spaces from our Campaign name – this’ll allow us to plug them in to the URL without any issues.
This is real deal, straight from one of our clients. For each of these keyword-level URLs, we had to generate the following: the base URL, the provincial subdomain that that campaign is targeting, the category ID (unique to the ad group), and a keyword-level identifier that feeds in to this client’s internal site search function.We also had to do this for major metro areas, too.
You *did* say that that number was oddly specific.
Spoiler alert: there totally is.
Spoiler alert: there totally is.
This tactic, as developed by Sam Owen of PPC Hero, makes use of your Impression Share reports to determine a maximum possible daily spend for each campaign. Then, using Excel’s Solver function, it solves for the maximum possible amount of conversions based off of your performance metrics. The wrinkle, in this case, is that I threw in both Google *and* Bing accounts to get an idea of the optimal budget allocation. Here, three Bing campaigns were recommended to get a big budget bump.
Now that I think about it, John Gagnon looks a lot like Zack Morris…
That Lost Impression Share report can include both Search and Display, if you want to get fancy.
Here’s all you need.
I’ve highlighted several campaigns that have a higher theoretical spend cap than our daily average. We have a branded search campaign AND a generic product campaign that have room to grow.
I’ve highlighted three campaigns worth mentioning: Our “Branded – Product 1” Campaign is recommended to max out it’s daily budget in pursuit of conversion volume. Our “Remarketing Lists” campaign gets the same recommendation.Our “Display – Keyword Targeting (NA)” campaign is recommended to be paused – based on the parameters that we’ve set for Solver. Parameters I’m going to show you now.
While intimidating to look at, it’s actually pretty simple. What we’re doing here is setting an objective – in this case, we’re looking to maximize the value that shows up in cell K45. This cell is the sum of all other cells in that column – it’s a conversion total. We’re then giving Solver parameters – the variable cells here are the “Solved Budget” column. We’ve giving Solver free reign to modify those values in pursuit of maximizing our objective. The constraints we’re putting in place say that our solved budget cannot exceed the maximum possibly daily spend. With this in place, it will then tell us exactly which campaigns should get our advertising dollars, based on our goal. You can do this for conversion volume, profit, revenue – anything.This gets really cool when you start working under a budget – you can dictate to solver that the total spend can’t exceed a certain amount. In this example, we’ve set a total spend cap of $600/day at the bottom of column H.
This CPA analysis looking at Hour of Day and Day of Week metrics? A Pivot Table.
And this lead-to-sale report directly tying our AdWords spend to closed leads from Salesforce? Came off a Pivot Table.
This analysis of the standard deviation of our Average Cost Per Click at a match type level? A PIVOT TABLE.
And this table that, when you rotate it, changes in size? Also a Pivot Table.