YT YouTube2Text

YouTube2Text → Transcripts → Practical Advice on How to Actually Make Smart Bidding Work for You | PPC TH 105

Practical Advice on How to Actually Make Smart Bidding Work for You | PPC TH 105 — Transcript

by PPC Town Hall · 9,498 words · 1,319 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00hello and welcome to another episode of
0:01PPC Town Hall my name is Fred vales I'm
0:04your host I'm also CEO and co-founder at
0:07Optimizer a PPC management software so
0:10for today's episode we have as guest
0:13Sean Walker from symphonic digital and
0:17he's been doing PPC for a long time and
0:19he figured out that the strategies that
0:21used to work in the past no longer work
0:24obviously we all know that but Sean went
0:27really deep on that and he figured out
0:29how does he get get smart bidding and
0:31automated bidding to work in today's
0:33framework where there's so much
0:34automation from the ad platforms like
0:36Google how many conversions does it take
0:38what bid strategies to run what to do if
0:41you don't have enough conversions so I'm
0:43really looking forward to hearing all of
0:45this very practical advice that I think
0:47will benefit all of us so with that
0:50let's get rolling with another episode
0:51of bpc town
0:56hall all right Sean welcome to the show
0:59good to have you on yeah good to be here
1:02hello Fred hello hello hey uh fellow
1:05Californian right this doesn't happen
1:07very often I guess maybe not California
1:11but I live in California so uh hey tell
1:13us where you live and uh what's your
1:15background yeah of course I'm in San
1:18Diego right now so we're in the good
1:20weather Place everyone gets jealous
1:22about um yeah I actually moved from
1:26Australia so I ran a performance team
1:29out there there for about three years
1:31and then before that was in LA but um my
1:36career started at Yahoo when Yahoo was
1:39actually cool now when you have a Yahoo
1:42email people kind of look at you funny
1:44but uh that's where I got my start so I
1:47was actually on the publisher side and
1:50then ended up working all through PPC so
1:54whether it's search social display um
1:58even worked at
2:00SEO as well and really my big background
2:04I I think I've done this probably for
2:05about 13 14 years going on but I'm a
2:10closet nerd I will say we live in a
2:13world where there's so much free
2:14education and I took advantage of that
2:17so got certified in Google analytics uh
2:20lots of coding platforms so JavaScript
2:23PHP Google scripts all that and any
2:26certification you can get your hands on
2:28I basically got my hands on so
2:30lots of Education there um but really I
2:33guess my bread and butter at symphonic
2:37is I'm the lead for strategy and product
2:40so when it comes down to putting
2:43strategy together I orchestrate the team
2:46I put together strategy with the team as
2:49well as provide technology and products
2:52for whether it's reporting or lead
2:55generation we' built even crms um at one
2:58small point in my life I actually built
3:01an onboarding platform um because I do a
3:03little bit of coding on the side to help
3:06bring new employees on so really into
3:09the tech part uh and we live in a really
3:11interesting world with AI and things
3:13like that now but uh that's that's a
3:16little bit of my my background yeah I
3:18know that's a fantastic background and I
3:20feel like uh in California either
3:23everybody has a script ready to go and
3:25sell it to Hollywood or everybody's
3:27building a piece of technology trying to
3:29uh be the next unicorn SAS company so
3:33sounds like you're just one of them um
3:35now also between Australia and San Diego
3:38do you like surfing or do you like the
3:39beach what's uh what's with those two
3:42places fishing and music that's my go-to
3:45so what's great is the ocean is 10
3:49minutes away from my house and then you
3:51have lakes that are 10 minutes away and
3:53mountains that are 10 minutes away so
3:55everything's really close um big bass
3:58fisherman so so I like to go out get
4:02away from the computer and go stare far
4:04away rather than up close that helps um
4:08and then I also play some music as well
4:10playing a Yacht Rock Band so that's
4:12pretty fun very cool well good it's
4:15great to uh to have you on the show so
4:18um there's actually a lot of topics I
4:20want to go talk a little bit more about
4:22like education and how you what kind of
4:24software you build and how you've been
4:25using GPT but let's get into the the key
4:28topic here today which was about bid
4:30management right and how do you outsmart
4:32Google's conversion based bidding
4:35algorithms and I think one of the first
4:37questions that most advertisers always
4:39have is how many conversions do I need
4:41before I can trust Google's algorithms
4:43to start doing a good job so uh can you
4:47talk a bit about those numbers yeah
4:49absolutely so Google is just following
4:52Suite
4:54of Statistics right so statistics says
4:58if you remember in school you need a
5:00sample size of 40 in order to get some
5:03kind of statistical significance and a
5:06reasonable level of accuracy Google
5:09brings this to 30 but I've seen this
5:11Jump Around um in some cases even 15 but
5:16uh I think the safe bet is 30
5:19conversions per 30 days per campaign
5:23that's the important part is even if you
5:25have 30 conversions in your whole
5:27account it doesn't matter it's all about
5:30each individual Campaign which uh I know
5:34we'll probably get into this but your
5:36structure is gonna change things right
5:39if you have 10 campaigns versus one then
5:42Google's going to have a harder time
5:43with the 10 campaigns versus one to try
5:46to reach that statistical significance
5:49right so it sounds like that need for a
5:51certain number of conversions plays a
5:54heavy factor in how you structure your
5:55account ultimately um now in terms of
5:59folks who have maybe much bigger
6:01accounts and they have hundreds of
6:03conversions or those that have much
6:05smaller accounts we'll talk about that
6:07as well but like how much sensitivity
6:09have you seen to this number of 30 like
6:12if if you go from 30 to 60 do you get
6:14like twice as good performance because
6:17Google has just that much more data or
6:19is that like a diminishing return like
6:21what have you seen there yeah uh this is
6:26such a big difference from when I
6:28started so so now we see a lot of
6:31sensitivity right so the more data you
6:34have the more accurate you're going to
6:36be but what we'll see is if you don't
6:39hit those 30 conversions in one campaign
6:42per 30
6:43days your campaign acts a bit wonky
6:47Google tries to spend as much as it can
6:51in the beginning when you're on let's
6:53say Max conversions or some kind of
6:56conversion based bidding strategy in
6:58order to learn and that learning phase
7:01can kind of kill some small
7:03businesses versus big you can handle it
7:06you got the budgets what we've seen with
7:08bigger businesses is if you're let's say
7:12you have you know a thousand conversions
7:14let's just go crazy the system does a
7:16pretty good job right so when you leave
7:20it on Max conversions or you do some
7:22kind of rows targeting if you're in
7:25e-commerce or cost per lead it does
7:27really well if you can get get there but
7:30the problem is we work with a lot of
7:32small to mediumsized businesses and it
7:35seems like the bigger advertisers we
7:37don't have a huge problem there we work
7:39on other things like ad copy and
7:42improving landing pages and things like
7:44that where Google does all the heavy
7:46lifting but the smaller
7:48advertisers it's a ton of work and it
7:51seems like it's just a struggle just to
7:53get to that point of statistical
7:56significance and trying to get purchased
7:59or or leads is all over the place one
8:03keyword that worked for you last week
8:05isn't the same keyword that works next
8:07week so we live in a day and age where I
8:11think signals are more important even
8:14than
8:15keywords and Google might even back that
8:17up they're they're leaning more towards
8:20targeting a lot of broad keywords if
8:22you've seen out in The Ether where
8:26people get kind of lazy I'll explain
8:27this so I went to a conference
8:30in Australia for YouTube and one thing
8:33that kind of stuck with me is people go
8:35to YouTube by typing you just you right
8:40and you'll notice any site you go to you
8:42might type the first three letters and
8:44just hope Google gets it right so almost
8:48don't want to say we're lazy but we're
8:50more efficient these days right so what
8:53ends up happening when you have
8:56something like this is things get a
8:58little bit more
9:00difficult to guess okay so people may
9:04not have the same signals they had
9:06before where they say hey I want to
9:07purchase this specific thing and their
9:09keyword is you know 20 years long um
9:13it's one keyword and you have to guess
9:14the intent of that and you can't guess
9:17intent on the keyword uh by itself if
9:20it's just one word you have to guess by
9:22signals like in Market signals um could
9:26be income could be the sites you browse
9:30um contextual places
9:32that there's probably a million signals
9:35that Google uses in order to find if
9:37someone's actually in Market you get the
9:40right person at the right time um rather
9:43than just solely relying on the keyword
9:45these days that's that's one of the
9:46challenges we've
9:48found yeah and that's interesting to
9:50sort of like restate what you're saying
9:52there is that historically the search
9:54engines just had less of a memory around
9:57who a person was and so was much more
10:00important to have a fully qualified
10:01keyword that exactly explained what you
10:03were looking for because that was the
10:05predominant signal to fetch the the
10:07correct and the most relevant ads but
10:10nowadays you're saying Google knows so
10:12much about you generative AI you're
10:14having conversations with these
10:15assistants that's basically serving kind
10:17of like your memory so now you can type
10:20something as simple as like three words
10:22the the three letters of the keyword
10:25that you were intending to type and the
10:26system already knows what you want and
10:28so it's able to pull in the right ads so
10:30you can no longer rely just on uh on
10:32that exact keyword it's it's about all
10:35of these signals that are auxiliary um
10:38which also maybe speaks a little bit to
10:39like upper funnel activity right if
10:41you're not present in those upper funnel
10:43stages where the memory um and the
10:45intent is building then it's going to be
10:47much more difficult for Google to say
10:49like oh you are the advertiser at that
10:51bottom of the funnel who is the right
10:53match for that thing that the user has
10:55been thinking about for two or 3 days
10:56and and gone on a complex consumer J me
11:01yeah yeah and we find that like back in
11:04the day you would just get these
11:05longtail keywords and they would work
11:07but now those don't seem to work as well
11:11as other keywords uh we're talking one
11:14to two keywords versus you know three to
11:17five let's say it it's it's backwards
11:21we're living in a backwards world where
11:23all these signals are more important
11:25than the keyword itself as to why that's
11:28the case
11:29I could only guess um I I think it comes
11:33down to we're trying to be efficient we
11:35expect things right away you know even
11:37when you're looking for a movie let's
11:39say on Netflix I mean you type in two
11:41letters and you hope the movie gets
11:43listed to the right you're not going to
11:45type the whole thing out so we're just
11:47moving that way I I I think there was a
11:50big conversation before about voice
11:53search where I I just don't hear that
11:56much Rumble about that anymore it used
11:59to be all about longtail keywords and
12:01and now Google has switched to Broad so
12:04there's something just changing out
12:08there to totally like huge changes right
12:10and I think when it comes to voice
12:11search the problem that I always thought
12:14was consumers want Variety in options
12:16and the problem with the voice search is
12:18that if you ask it whatever it is it's
12:21really difficult to give it three to
12:23five options like you would see on the
12:25search results page where you can
12:26visually very quickly take that in and
12:28go to the one you want voice search you
12:29basically have to wait for the first one
12:31to be read to even be exposed to what
12:33that second one and the third one is and
12:35people I mean it's all about shortcuts
12:37right people don't want to spend more
12:38time doing things they want to spend
12:39less time doing things but I think where
12:42generative AI is changing that game now
12:44for voice search is that we no longer
12:46have to think about like keyword to
12:49result it's actually about it's a
12:52conversation I can explain to the
12:54assistant what it is that I want and
12:57it's it doesn't have to give me three
12:59options they can say okay well based on
13:00what you've explained here's a
13:01summarization of the options that you
13:03have and as you start to like understand
13:05what it is that you actually want it
13:07leads you down to that one answer and so
13:10I think voice search and voice uh based
13:13advertising or chat based advertising
13:15obviously is going to come back in in
13:17full force just because we have
13:18different and better technology that we
13:20just didn't have in the past um but like
13:22you're saying that means it's no longer
13:24about one keyword to a bunch of ads it's
13:27about you know many conversations
13:29that consisted of thousands of words of
13:32actually talking to your Bot and that's
13:34the thing that creates the signal that
13:36Google Google or chat GPT or whoever is
13:39able to uh to use to say here's the ad
13:42and then the other thing that's
13:43fascinating about that is there's some
13:45notion that your assistant or Google
13:49actually knows before you've even
13:50started typing anything what it is that
13:52you likely and you know you think about
13:54your phone right like you open the app
13:56to find the search screen on your phone
13:58and it already has like the 10 most
14:00likely things you're about to do right
14:02now because it looks at the context of
14:03your day your location what you
14:05typically do um and Google is is much
14:08similar in that regard and so may maybe
14:11rather than them just showing us some
14:12stuff that they already knew we were
14:14going to search for they're like yeah
14:15let's wait let's let's have them type in
14:16three characters so they at least feel
14:18some sense of control uh but it is
14:20pretty crazy what the technology can
14:21actually do these days yeah and you know
14:24I'll say one thing I feel like I'm
14:26trying to be a little bit of a hero for
14:28the small guy here
14:31so let's say five years ago even this is
14:35all great right but now Google took away
14:38from us
14:40unfortunately targeting these longtail
14:43keywords so we we write in these long
14:46prompts right or you'll go to Reddit and
14:49type in this long thing uh so these
14:51searches do exist but the problem is is
14:54if you try to buy these Search terms
14:56Google will say to low volume
14:59right and then you can't actually use it
15:02you can't buy it just you just won't
15:04show you can buy the keyword but you're
15:06just not going to show for it uh and
15:08then also unfortunately in your search
15:10query report you don't get to see all
15:13the terms I forget the percentage but
15:15it's a very low percentage of what you
15:17actually get to see so if that came back
15:22and if Google's listening bring it back
15:25but if that came back with all this AI
15:30prompting I mean that would be a great
15:32marriage but unfortunately the little
15:34guy is stuck right you can't buy those
15:37terms that you know are going to work um
15:40or very pointy you have to you have to
15:43go Broad and you have to spend those
15:46dollars because you're just not gonna
15:48you're not going to show up Google will
15:49literally flag you and say too low
15:52volume and your keyword won't show um
15:55that's been going on for years and years
15:56and years and that's where we kind of
15:59struggle with some clients I will say
16:02because uh a lot of people live in the
16:04past uh and they have terms in their
16:08head that they know people search and
16:10they say let's just buy these and then
16:12they we don't even show for them um
16:16that's been a big struggle for us with
16:18small businesses and especially with
16:21singles just going back to bidding um
16:26the problem is businesses small
16:29businesses have a higher cost to operate
16:32per unit right if you're a big business
16:35and you you bought a big building and
16:38you fill it up with employees you're
16:39you're at economies of scale where to
16:42produce one unit it's going to be a lot
16:43cheaper but if you're a
16:45startup that unit that you're selling is
16:48going to be so expensive so your targets
16:51are actually harder to hit you might be
16:53able to operate at a 2 to one uh as a
16:57huge business right getting rid of
17:00operational costs but as a small
17:01business we see often targets where
17:04people say oh we got to hit a 5 to one
17:06or a seven to one which is that much
17:08harder because you're competing with
17:10these big people right you're competing
17:13um we all have the same car these days
17:16with AIS and algorithms but the
17:19challenge is how do we drive it
17:22differently you know if you're a small
17:24business how do you how do I drive it
17:25differently than the big business um
17:28that's a struggle and I'm sure any small
17:30businesses end up watching this you know
17:34probably the best piece of advice would
17:36be it things don't work how they used to
17:39anymore so you're supposed to simplify
17:41and one of the key things that I keep
17:43hearing there is instead of having
17:45thousands of campaigns and tens of
17:47thousands of AD groups and like these
17:49really complex structures that go after
17:50the long taale it's really much more
17:53about simplifying it in campaign
17:55structure just have a handful of
17:57campaigns and the campaign is really
17:58focused on what is your budget or what's
18:00the Geo Target or what's the specific
18:02type of conversion you're trying to get
18:04for that campaign and then you match
18:06that up with some of the automations
18:07from Google like the broad match keyword
18:10responsive search ads automated bidding
18:12you put those three together um
18:15presumably Something Good Can Happen but
18:16that good thing can only happen if you
18:18give it the key signal and this is what
18:20you were talking about right the key
18:21signal to make all this automation work
18:23for you whether you're a big guy or a
18:25small
18:26Advertiser um is that conversion data
18:28and so ideally you have at least 30
18:30conversions per 30 days per campaign but
18:33tell us what do you do if you're one of
18:35these smaller advertisers and you can't
18:38get through those 30 is there some other
18:40Benchmark some other signal that you can
18:42push into Google to get things maybe
18:44rolling in the right
18:45direction yeah so this is probably the
18:48most important part of the conversation
18:50I would say
18:51so most small businesses are not going
18:55to hit that so let's think of a scenario
18:59if you if let's just say you have a
19:02$2,000 budget right and your cpc's are
19:06$5 which isn't crazy that's normal and
19:10you have an e-commerce product where you
19:14have a conversion rate of 10% right so
19:17you end up I did the math right I think
19:19you get about 40 purchases okay would
19:22you would say great that's 40 signals to
19:25my conversions but if that's spread
19:28across 10 campaigns actually even if
19:31that's spread ac across two campaigns
19:33you have not hit the threshold that
19:35Google needs and for some small
19:38businesses that's an impossible feat to
19:40get to so the question is what the heck
19:43can I do to beat this um there's a
19:46couple things so you just have to think
19:50simple right what's the next thing above
19:52a purchase or let's talk B2B and
19:55e-commerce at the same time so above
19:58purchase would be add to cart or for B2B
20:02lead generation let's
20:04say if you can't get somebody to fill
20:06out a form what what's something else
20:09that you can get that's right before
20:11that I kind of see the order of
20:13operation going like this um you see an
20:16ad that's your first signal your second
20:19signal is you click the ad then you go
20:21to the website if you make it there
20:23because some people drop off and then
20:25when you're on the website you do
20:28something you engage right you either
20:29scroll down the page you read some
20:32content you go to another page past that
20:36it's more engaging content it it might
20:39be a gated asset or or even ungated
20:42asset where you open up pdf or something
20:44like that um and you see where I'm going
20:47with this so there's all these
20:49conversions that you could build out
20:51from the point of someone clicking your
20:53ad visiting your website all the way to
20:55purchase and lead and I'll just give
20:57some examples of things that we've built
20:59out that has worked for us with small
21:03businesses so one thing that we use is
21:07this conversion called engaged
21:10visitor and it's pretty interesting um
21:14it does two things so you go to the
21:17website and if you scroll down about 50%
21:20of the page that would be a normal
21:23conversion that you could launch however
21:26if you come back a hundred times let's
21:28say say it's going to fire 100
21:30conversions and you might not want that
21:32but we built a cookie based conversion
21:36that only fires per unique user so if
21:40you go to the site you scroll down 50%
21:42of the way a cookie will fire and it'll
21:45say all right Fred's been to this site
21:48he's engaged with the site he's not a
21:51bot um he's a higher quality user right
21:55and if you come back a 100 more times
21:57it's still just going to count you as
21:58one so we've been using this signal for
22:02a lot of B2B we've been using it for a
22:04lot of e-commerce just to have something
22:07that's going to get rid of the Bots and
22:09to get people who aren't going to go to
22:11the site and be idle so that's been
22:13pretty successful and I think we have it
22:16across most of our clients now um takes
22:19a little bit more scripting I will say
22:21you can you can get it done in Google
22:23tag manager where you have to have some
22:25kind of script that can set a cookie
22:28once you scroll down the page and then
22:30also read the cookie and say okay Fred's
22:33been here before he's already been
22:35tagged don't fire this conversion again
22:37if someone wants to know more about that
22:40reach out to me on LinkedIn I could
22:42probably give my my Link at the end of
22:44this but um that's one good thing the
22:48other thing I will say
22:50is product views for e-commerce that's a
22:53big one if someone is looking at a
22:55product and I don't mean you landed on
22:58the product page I mean you went to the
23:01website you ended up going and looking
23:02at a product um product remarketing is a
23:08good signal right so uh for B2B I will
23:14say any kind of content clicks or
23:17outbound emails you know where you're
23:20not necessarily giving a lead but you're
23:22showing intent to contact anything
23:25that's a little bit softer so I guess
23:27the summ of what I'm saying is find a
23:30soft conversion where you're going to
23:32get that volume that can be fed into the
23:35system and give you a good indication
23:37that this person is much more likely to
23:42convert to something else like a lead or
23:44a purchase versus someone who just went
23:46to the website and bounced right away
23:48maybe even spent two seconds on the site
23:50or it's a bot because we have a bot
23:52problem these days so every Advertiser I
23:56would recommend if you're small you have
23:58to set up these soft conversions and
24:02don't assume that you're just going to
24:04get a flood of purchases or a flood of
24:06leads because we're just not seeing that
24:09these days especially because you need
24:11so much budget to get that statistical
24:14significance right so think about it as
24:17a hierarchy of what is it that the
24:19typical consumer goes and does as they
24:21go to the ultimate conversion which
24:22would be the sale um and and both in
24:25e-commerce and Legion I think the
24:26ultimate conversion is the sale is
24:28closing the deal um but then the beauty
24:31is that if you think about that funnel
24:33there's different stages right so in
24:35what is your approach for not
24:36necessarily saying okay I have to pick
24:38this third stage in The Funnel of
24:42desirability but do you create
24:44conversion actions for each of those
24:45steps and then sort of toggle those on
24:47and off as you start seeing there's
24:49enough volume in that more desirable
24:52action so let's get more specific let's
24:54feed more specific data to the ad
24:56platform so it can bid for the thing
24:58that's a little bit closer to what I
24:59really want or how do you manage that
25:02whole um like dealing with shifts and
25:05volume at the different conversion
25:06action levels yeah so we have a specific
25:10approach so if it's a if it's a big
25:13client and we have budgets what we'll do
25:16is the same thing as we do for a small
25:18client we'll still build out those soft
25:20conversions
25:22and the the hard conversions I will say
25:24which are purchases and leads um and
25:28we'll start off with the thing that's
25:30more difficult to get and we will at
25:32least try Max conversions we'll try the
25:36cost per if we have um a goal right and
25:40if we don't get there we have our life
25:42raft we have our backup plan right so we
25:46do build out actions for all these
25:48before we even
25:50start and some campaigns might get there
25:54quicker so if you're just using a
25:57setting per campaign that just goes with
26:00the full
26:02account I almost don't want to recommend
26:04that because each campaign might be
26:07different one campaign might be doing
26:09better than the other where the other
26:11one never gets that statistical
26:13significance so what you can do is just
26:15pull in that softer action right and
26:19when it gets to the point where you can
26:20actually get a harder lead you can uh
26:23you can switch it and and pause on that
26:25for a second so you said something Ong
26:28the lines of account level Campaign
26:30level so I think you're talking about
26:32the different levels at which a
26:34conversion action can exist can you go a
26:35little bit deeper on um explaining that
26:38for folks who may not have necessarily
26:40used that capability yeah absolutely so
26:44by default when you set up conversions
26:47Google will put this as a primary target
26:51right you have to manually go in there
26:53and set it to secondary and the
26:55difference between the two is your
26:57primary
26:58is going to be included in your
27:01optimization settings where if you just
27:03start a campaign um and you choose
27:07account level it's going to pull in all
27:09your primary assignments by default but
27:13you can also use the secondary ones if
27:16you want so you can pick and choose what
27:18you want um but by default you're going
27:22to use the account level however if you
27:24use the campaign level which you're
27:26allowed to you choose whatever you want
27:29in order to optimize too because I I
27:32think every campaign is going to be
27:34treated a bit different um especially
27:36like let's say you're a big
27:39brand and you're running non-brand and
27:42brand maybe your
27:44brand is doing much better than your
27:47non-brand you might have different
27:49strategies you might have different
27:51conversion points per those so using
27:54that account level may not be the right
27:56approach might be better to use the
27:57campaign level
27:59right okay so you have these different
28:00levels of conversion actions you can
28:01make them primary or secondary and
28:04basically you can make it primary as
28:05there's enough volume to say this is the
28:07signal that you now need to use Google
28:09for my bid
28:11optimization um and maybe make some of
28:13the old primary secondary so they're no
28:15longer considered because they might not
28:17be as desirable maybe they're upper
28:19funnel micro conversions so that that
28:22sounds like one approach right is um use
28:24micro conversions or softer conversions
28:27to guide the system system there's
28:29another approach that I think you're
28:30advocating which is to do with account
28:33structures um so talk a little bit about
28:35how you think about the number of
28:37campaigns you have um and maybe as
28:40opposed to making more campaigns
28:42building more ad groups how should
28:43someone think about that yeah absolutely
28:46so I came from the world of working in
28:48automotive for search so I thought every
28:52account was going to be like this right
28:53oh you get $10 million and you can build
28:56whatever the heck you want
28:58not the world we live in and especially
29:00not the world we live in with small and
29:02medium advertisers so now what you want
29:06to do is get some idea if you have an
29:09analytics platform let's say using Adobe
29:11or or Google you can get some idea of
29:14what your conversion rate is uh for for
29:17something whether it's direct or organic
29:20you you start somewhere and you get an
29:22idea of this is my conversion rate and
29:25typically advertising is not going to be
29:27as good is someone who just goes
29:29directly to your website but you can get
29:31some idea to where you can reverse
29:34engineer and say I'm going to need this
29:36many
29:37leads or sorry this many clicks visitors
29:41to my website to get a lead then you
29:43could back into your budget right and if
29:47you find out that you know what it's
29:50gonna take me $100,000 to get
29:53statistical
29:55significance uh in that scenario it's
29:58not a great idea to break out into a ton
30:01of campaigns right because Google's
30:03going to look at each campaign and say
30:06are you reaching that statistical
30:08significance now if you're in a case
30:11where you can reach that and do so with
30:16like 10 different campaigns if it's
30:19advantageous for you to split out themes
30:22and you can still reach that statistical
30:24significance that's fine but I would say
30:26if you had to choose one priority is
30:30have a structure that is in line with
30:33getting enough conversions so um every
30:37account that we do now starts simple and
30:40we make it more robust over time versus
30:44back in the day we used to build out
30:45these giant campaigns um this huge
30:49account structure and then dial it back
30:52we now you want to do the opposite start
30:55small and then build as you see where
30:58there's you can build
31:00on that that's fascinating so you start
31:02small and then you expand from there
31:04there's a different approach to
31:05achieving sort of the same thing which
31:07is more narrow targeting right so I
31:10think most advertisers they come in and
31:11they say hey we're ready for a lot of
31:13business like send it my way um but is
31:17it sometimes better to take a more
31:19narrow approach and maybe do some geot
31:21targeting and do some exclusions and
31:24build from there or do you tend to not
31:26put too many restriction in place from
31:28the GetGo yeah
31:31so we we do want to put restrictions in
31:34so say you've got to the point where
31:37let's just say you even have one
31:39campaign you said this is as much as I
31:41can do uh I only get a handful of
31:43conversions I only spend $500 a month
31:46let's just say in that case you need to
31:49make a decision on feeding signals that
31:54you know of beforehand so Google
31:57Advocates that let us do the work right
32:01let us spend a bunch of money we'll find
32:04your people and we'll just keep going
32:07however a lot of people don't have uh
32:11the budget to do that to start right so
32:13what you have to do is use whatever data
32:16you have your marketing data consumer
32:18surveys um you know talking to people
32:22email you know the power of
32:24communication is Big um and you find out
32:27where people are so we've seen a couple
32:29things a different couple signals um
32:33let's talk about Geo real quick there's
32:34City and Rural right if you have a
32:39business that's going to thrive in the
32:42city um do City targeting if you're
32:44going to have a business that's going to
32:46thrive rural then Target that right
32:49different states perform differently I
32:52will say this a lot of people put things
32:54on observation but if you know that
32:58let's say Texas versus New York let's
33:01say Texas um as far as your your
33:05clientel goes has double the conversion
33:08rate as anywhere else start there if
33:11your if your budget can handle it just
33:13start there you know pick the lwh
33:16hanging fruit as people say because you
33:18know that there's more converters there
33:20and you could put that on explicit
33:22targeting rather rather than observation
33:25where you're still targeting everyone
33:27and you're learning
33:28if you already know Texas does well
33:30Target Texas I will say this as well
33:33there's other signals like if you're
33:35selling a high-end brand let's say
33:39where you know it's ,000 to purchase
33:42something or more you might want to just
33:44explicitly Target the top 10% of income
33:49earners which is a signal in Google
33:52rather than targeting everyone or what
33:54you can do because what we find is the
33:58unknown when Google doesn't know what
34:00income bracket you're in um that tends
34:02to be a big one that does well but you
34:04can do exclusion so say you know that
34:07the lower 50% income is never going to
34:10buy your product just start off
34:12excluding that don't don't guess if you
34:14already know just start there and then
34:17other things like I'll just give an
34:19example of one client we have a client
34:21that works in um in sports right so
34:26baseball volleyball things like that
34:28you can choose in Market targeting
34:32explicitly where you can find people are
34:35just interested in something and maybe
34:39maybe what happens is you grow you start
34:42by just targeting let's say baseball
34:44right people are interested in baseball
34:46or baseball equipment and you find that
34:50you start scaling you could then put
34:52that on observation and go after other
34:54people but in the beginning I think the
34:57best way to approach your budget is do
35:00the best targeting you can in the
35:02beginning and see how far you can go
35:04with the lwh hanging fruit and then
35:06expand from there but at least you know
35:09how well you can do right from thego uh
35:12versus the other way around of guessing
35:14and spending a lot of money to
35:16guess yeah and this is interesting
35:18because you're you have competing
35:19objectives right on the one hand you're
35:21saying you need to have around 30
35:23conversions per 30 days but on the other
35:26hand you want to Target to the places
35:28that perform better such as in your
35:30example Texas having twice the
35:31conversion rate or maybe you want to be
35:33in market for a certain audience but
35:35that obviously reduces the number of
35:37conversions that you get um but so I
35:40think once you put those two together
35:42what it sounds like you're saying is
35:44Target specifically where you know you
35:45will perform best and then fix the
35:48conversion issue as a second step by
35:51going maybe higher up in the funnel and
35:53not looking at the sale but looking at a
35:54soft conversion event that precedes the
35:57the actual thing that you want it sounds
35:59like that is the better strategy as
36:01opposed to going really broad getting
36:02lots of volume but maybe the wrong kind
36:05of
36:05volume right and go broad where Google
36:09says you need to go broad so you might
36:11have to do broad keyword targeting
36:13because we've seen that it doesn't work
36:16otherwise if you get too tight but go
36:20broad there and tighten up your
36:21targeting like you're saying your Geo
36:23tighten up your in market segments
36:25things like that and then with that same
36:29approach start with your your hard
36:33conversions you'll see in the first week
36:35right if that's not working then all of
36:37a sudden you have to bring in your soft
36:39conversions and feed it some kind of
36:42data to where it can go after more
36:45quality users and you could eventually
36:47get to your sought after goal which is
36:50you know sales um and I'll mention one
36:53thing with B2B what I I find interesting
36:55is there's all these systems like
36:57Salesforce and HubSpot we can have
37:00marketing qualified leads and
37:02opportunities and we always connect
37:04these systems but I'll say even for
37:06large clients there's never enough
37:08volume on these to actually optimize to
37:12those there they they seem to be like a
37:14nice to have in the system to say oh I
37:16eventually got to this um but the
37:19problem is you just never get that St uh
37:23statistical significance right well what
37:26you're describing is like you have the
37:28funnel and we've talked about like the
37:29lead is what you kind of want but
37:31there's all these things that come
37:32before it when you go into the CRM it's
37:34actually even beyond the lead it's
37:35deeper down the funnel of higher levels
37:37of qualification so the the volume of
37:41data becomes even more sparse um so it's
37:44a nice problem to have I mean obviously
37:45if you spend a million dollars a month
37:47on ad ads you probably do have enough
37:49CRM data to feed the system um with
37:52those mqls and sqls but uh it's if
37:55you're a smaller Advertiser then
37:58CRM sounds like a total luxury but uh
38:01not something that's immediately going
38:02to give you some gains yeah like the
38:06problem is you're trying to predict
38:07Norms right you might have if you're a
38:10smaller client let's say you have uh
38:14five sales in Salesforce which might
38:17sound like a small number but it's it's
38:19not in the B2B world for a smaller
38:22client and you might see that one person
38:25is in we'll just go back to my example
38:28New York and one person's in Texas and
38:32they both like
38:34sports you cannot use a sample size of
38:36five to say all right we're just going
38:38to explicitly Target Sports and we're
38:40just going to Target people in New York
38:42and Texas this is where I have a problem
38:44with some of the data that people lean
38:47on um CRM data to make decisions I think
38:51you always have to go back to that
38:53statistics class and say do I have
38:56enough samp
38:58to predict Norms right is is this going
39:01to be the norm or is somebody going to
39:03squeak in there from Florida you know or
39:06Washington and then throw your plan out
39:09the
39:10window yeah and think about the agency
39:13perspective like you have to teach your
39:15clients to have these Fair expectations
39:18because if they're going to be basing
39:20their judgment on five conversions in a
39:22month while guess what next month you
39:24only got four conversions which really
39:26that was a crap right that could have
39:27happened either way that's outside of
39:30the the statistical range but they're
39:33going to look at that and be like oh my
39:34God Sean why did you drop 20% of
39:36conversions it's like well next month
39:38we'll have 20% more because we'll
39:40randomly have six instead of five but
39:42that's not an actual statistical pattern
39:44that we can do anything with right so
39:47you're setting up I think you're setting
39:48yourself up for a lot of hard
39:50conversations if you allow your customer
39:52to to look at those small numbers as the
39:55uh the absolute truth yeah they're
39:57they're nice to have but I wouldn't make
40:00any big decisions based off of that yeah
40:03so I mean obviously at some point
40:06something happens automated smart
40:08bidding may not
40:10work whether it's due to data being
40:12broken maybe the volume just isn't there
40:14no matter what you do what's kind of a
40:16backup is there is there some level of
40:18automation you can still put in place
40:20that you control based off of the the
40:22type of signal you have yeah uh so a lot
40:26of the times if we're starting a brand
40:29new campaign where there's not much
40:32background and the client doesn't really
40:35know you know exactly who the audience
40:38is we will start off with manual
40:41bidding right just to get some volume
40:44out there but you can still make
40:46Intelligent Decisions so outside of
40:49adjusting ad copying keywords and
40:51negative keyword the the obvious stuff
40:53you know if someone if if you're not
40:55handing out discounts and people are
40:57typing in your brand name and discounts
40:58you probably want to negative out
41:01discounts things like that um there's
41:04the obvious stuff but then there's the
41:06other things that are how can I optimize
41:10when I have limited conversion volume I
41:12think is what we're saying so something
41:15is better than nothing so why while you
41:18can't make big decisions off of what
41:21you're doing you can do a couple things
41:23um we built out some scripts that do
41:27some of the lifting that Google would do
41:29in their automation I will say this we
41:31don't have big decimators right like
41:34they don't go in and just start pausing
41:36everything um it's very slight changes
41:39so we have a couple things like device
41:42automation so even if you have five
41:45conversions what you can do is just have
41:48a level setting uh where you can control
41:51in the script where you bid up where you
41:53see performance and bid down where you
41:56don't based off of conversion rate with
41:59limited signals right but you do it in a
42:02way where you're not changing bids by 50
42:04or 60% and you're not decimating the
42:08campaign you're doing something which is
42:11better than nothing so even if it's a a
42:135% change and you're testing a theory
42:17you're doing so you're doing what you
42:19would normally do as a human but you're
42:21letting a robot do it so this is this is
42:24pure automation um
42:27there's also in market segments uh we
42:30have a script for that that looks at
42:32segments and does the same thing it
42:34would for device same thing for GEOS I
42:37would say if clients don't know exactly
42:40where their customer base is I think
42:42it's good to start off by just uploading
42:4550 States on observation so you can make
42:49changes to them apply bid multipliers so
42:52we have a script that will make bid
42:54multipliers work there's also
42:57a campaign script that we have where if
43:01we see a campaign doing well a human
43:04would go in there and up the budget and
43:06decrease the budget on the other
43:07campaign uh we have robots that'll do
43:10that um so you're kind of seeing a theme
43:13here it's basically anything we would do
43:15as a human we're letting a robot do and
43:18then there's also a couple uh good ones
43:21we have here is a schedule Creator so
43:23you hit a button it creates your Monday
43:25through Friday uh sorry Monday through
43:29Sunday where not only do you get to
43:32automatically create a schedule but you
43:34get to add bid multipliers to that as
43:36well and just the last couple things is
43:39if you're limited by budget from your
43:43keywords there's one quick fix there so
43:47the Quick Fix is if you have the budget
43:50and a campaign is doing well then put
43:51more budget into it but if you don't and
43:54you want to get rid of that message what
43:56you can do is actually just decrease the
43:58bids of keywords within that campaign
44:01because essentially what you're doing is
44:03you're paying more than you should for
44:05your keyword and it's stopping a partial
44:08way of the day so Google saying you're
44:10limited by your budget we're only going
44:11to show you for a portion of the day so
44:14you can actually get more clicks by just
44:16decreasing your bid and we have a
44:18script where it will decrease bids for
44:22keywords that are not converting until
44:24that message goes away um I will say
44:28this is not perfect but it's better than
44:31nothing right it the fear is in some of
44:36these if you think a keyword is going to
44:40do well for some reason it's done well
44:42in the past um could this bid it down
44:45because it h hasn't had a conversion yet
44:48yes so there is a caveat there if you
44:51turn all these on and there's 10 things
44:53bidding down or up could things get
44:55expensive yes so you have to keep an eye
44:58on it it's not robots taking over the
45:00world it's more you need to be there you
45:03need to wield the hammer and make sure
45:05that it's working correctly and keep an
45:07eye on it
45:08right yeah absolutely and instead of
45:11spending manual time doing these
45:13adjustments that you would have done
45:14anyway get back your time by scripting
45:17it or or you could use Optimizer and
45:19Rule engine can do that for you without
45:21coding use these tools and Technologies
45:23so that you can take back your time and
45:25actually have more of a strategic
45:26thought about what else should happen in
45:28that campaign because that's really the
45:30human value at this point I think right
45:32um so it sounds like you're lever
45:34leveraging a lot of cool scripts to give
45:36you back time now as far as like taking
45:39back time um you alluded to some gen in
45:42the beginning of the show but uh any
45:44cool tools that you're using there
45:46Beyond uh scripts or are you even
45:48building scripts with GPT like what are
45:50you using Genai for that saves time in
45:52an agency setting yeah so um I will say
45:57that the coding has helped if if anyone
46:00can stomach it you know and sit there
46:02and and stare at a screen for a while um
46:06scripting does help just being able to
46:08read it you know and being able to write
46:09it because then you could create these
46:12things these tools almost like widgets
46:15right um leave the Enterprise stuff to
46:18to you Fred but you know the little
46:20things um being Scrappy I I think it has
46:24helped so much and as kind of part of
46:27that we use uh writer. me I don't have
46:31you heard of that
46:33before
46:35um I don't think I know that one okay I
46:38mean there's so many of these things so
46:40it's it's it's like a version of chat
46:42GPT but I will say one thing is when we
46:46were vetting a couple things this one is
46:50a super
46:52affordable and it does a really good job
46:56with search ad copy uh we had Jasper
47:00chat GPT and a couple other things and
47:04we found that this one was just kind of
47:06on the money with the ad copy I was
47:09actually wowed from it so we ended up
47:11using it this helps not only write ad
47:14copy as a start so let's say you have
47:18you know 20 ad groups you need to write
47:20for this is a good start for it where
47:22you can go in there and then tweak it um
47:25see if it makes sense um starting
47:27keyword development I know we all end up
47:29in the Google ads tool but this I always
47:33have it open I'll just put in you know
47:36the brand name and'll spit out 10
47:38keywords that I can copy and then put in
47:41uh Google ads and then all of a sudden
47:43it'll give me just more robust keyword
47:46examples well one thing I like about
47:48these generative systems like the
47:49keyword generation and the attex
47:50generation is that you can start to
47:52prompt with like put yourselves in the
47:53shoes of and then you put in your target
47:56audience
47:57and the typ of keywords and the types of
47:59messaging that it comes up with in the
48:00ad is like it's really specific to that
48:03scenario so I was recently testing this
48:05out for uh not translating but
48:08localizing an ad so taking an English ad
48:11and instead of straight up translating
48:13it to French I said could you rewrite
48:15this ad so that the French user would
48:17find it more appealing and so it starts
48:19making word selections that keep the
48:21core message the same but sort of the
48:23Nuance of the the UVS that it's putting
48:26in are highly targeted to a French
48:28audience cares about like the the
48:30Elegance of something as opposed to the
48:32the freshness of something which the
48:34Dutch care much more about and so the
48:36types of ads that come out actually fit
48:37those audiences and and that's the
48:39brilliant thing that GPT can do that so
48:41many of the old school tools just
48:43weren't able to yeah I like that and I
48:46think um I think overall you know I I
48:49don't want to speak for everyone but I
48:51think the right way to use these things
48:53is just to speed up your workflow I
48:55don't think it's supposed to replace us
48:58eventually you get there right you
49:00eventually get to write your ad copy but
49:02instead of staring at the screen and
49:04doing a ton of
49:06research let these tools do this for you
49:09you know and you know fact check of
49:11course but I think it's supposed to
49:14speed up workflow so that's that's why I
49:16use these things um you know couple
49:21simple things too I have it open just
49:23all day long just to even answer simple
49:26questions
49:27you know which is weird because I used
49:29to do that in Google but
49:33um you know I can't even say why I think
49:36I have this open so much now that I'm
49:38finding myself prompting more though
49:41more than when I go into Google uh just
49:44because I do a lot of things there um
49:48schedu I got to tell you A cool one that
49:50we did too
49:53um one thing I was finding myself doing
49:56every sing single day people go are you
49:59available like send me your your
50:01schedule or whatever you can't always go
50:04in and share your calendar with every
50:06single person you know you might be
50:07talking with reps and um clients and
50:11different things and you have to go into
50:12your calendar look go in there type it
50:16out and this was just becoming too
50:19cumbersome so with some scripts in
50:23Google Sheets uh I wrote this thing
50:25where you you connect your
50:28calendar and now I just put in the start
50:32and end date of like what date I'm
50:34interested in and I hit a button and
50:37it'll just spit out the days that I'm
50:40not available and the times and I copy
50:44this format and it takes two seconds I
50:46copy this I throw it into writer uh Ai
50:51and it spits out an email that's like
50:54like I wrote it you know and it says I'm
50:56not a available at these times and I got
50:58to tell you this has saved me I don't
51:01know maybe like two hours a week because
51:04I'm I was doing this all day long and
51:06I'm like surely there's an answer for
51:08this and I went on Reddit and all these
51:11things nobody really had anything like
51:13this so I just kind of hacked it and I
51:15gotta say it's one of those dumb things
51:17like that such a little thing but it
51:20added up so much time that that I was
51:23like you know what we got to do
51:25something here
51:27I think a lot of people might pay you uh
51:29you know three5 doll a month just to
51:30have that little capability maybe that's
51:32a business right there I know I would
51:34pay myself it's
51:35helped very cool um well awesome so yeah
51:40tons of good advice there as far as how
51:42to make bid management work especially
51:45if you're a smaller Advertiser so thanks
51:47for all of those tips uh any any final
51:50takeaways there that you think people
51:52should know I think there's one thing I
51:54missed I know you're saying how do you
51:56structure you know campaigns and AD
51:57groups I will say I think we live in the
52:00day where it's better to separate ad
52:04groups and it is to separate campaigns
52:07because all
52:09the optimization is at the campaign
52:12level so if you're sitting there making
52:13decision you know I want to break out my
52:16themes and as long as you don't need to
52:18control specific budget just break it
52:21out in the ad group rather than breaking
52:24it out at the campaign level I think you
52:25should only really breakout campaigns if
52:29it's
52:30truly a performing completely different
52:34than other campaigns if it's brand
52:36versus non-brand for example uh if you
52:39have to have a dedicated budget for it
52:42and if the like targeting is different
52:45because you can't Target different
52:47locations and in a different ad group
52:49let's say so certain parameters you have
52:52to use campaigns for but if if you can
52:54avoid it just do it at the ad group
52:57level and let Google use that
52:59statistical significance that will come
53:01with the reduction of campaigns yeah
53:03absolutely like understand why these
53:05structures exist what their capabilities
53:07are and if you need to make a division
53:09based on one of those specific
53:11capabilities then it usually make sense
53:12to build a new campaign um but if you're
53:15going to have multiple campaigns with
53:17the same budget the same Geo Target the
53:19same day parting then like you're saying
53:22you're maybe just making more
53:23complication for yourself than is
53:24necessary you're also going to confuse
53:26the machine give it less data don't use
53:29ad groups instead yeah and I would say
53:32um just to recap if I was goingon to say
53:35like three
53:37things be prepared for little conversion
53:42volume right and I know we don't like to
53:44hear that but be prepared for it build
53:46those soft conversions from the start
53:48right away out of the gate and start
53:51with the smaller structure if there's
53:54three things to remember I would say out
53:56of all this those are the three things I
53:58would I would give away okay well that's
54:00awesome thank you
54:01Sean if people want to get in touch with
54:04you or symphonic where should they go so
54:08symphonic
54:10digital.com I know there there should be
54:12an About Us section where you have a
54:15smiley face of me where you can get a
54:17hold of me on
54:18LinkedIn or I don't know if I can give
54:21you a link I we'll put it in the show
54:23not so we'll put all the links in the
54:24show notes but uh reach out to symphonic
54:27digital and Sean is also like everyone
54:30on LinkedIn good place to get a hold of
54:32him so uh yeah everybody thanks for
54:34watching if you've enjoyed this episode
54:36and want to see more of them please
54:38subscribe and leave us a review as well
54:40and we'll be back with the next episode
54:42of PPC Town Hall soon Sean thank you so
54:44much for joining and everyone thanks for
54:46listening or watching

About this transcript

This page contains the full transcript of Practical Advice on How to Actually Make Smart Bidding Work for You | PPC TH 105 by PPC Town Hall, generated from the public captions YouTube serves with the video. The transcript has 9,498 words across 1,319 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.