| This 47 message thread spans 2 pages: 47 (  2 ) > > || |
|PPC Profit Maximizing (not ROI targeting) SEM bid management tool |
| 10:50 pm on Dec 5, 2007 (gmt 0)|
Hey everyone, Iíve been researching bid management agencies and I was looking for feedback and reviews. Iíve been disappointed in what Iíve seen so far and Iím debating whether to have my programmers build an in house solution. Thatís not my #1 choice because I would think that these companies that pour millions into their solutions should be able to crush what we put together.
Iíve read other threads on webmasterworld about this topic, but I did not find the answers I was looking for. Iíd greatly appreciate any insight and reviews you might have about SEM agencies.
Hereís what Iím looking for:
(1) A profit maximizing algorithm (or at least something smarter than account level ROI targeting). ROI and profit maximizing is not the same thing. For example, I could minimum bid and watch traffic trickle in while I bank $30 of profit at 900% ROI. Iíd rather bid higher and increase my profit at the expense of a lower ROI. But at what ROI point does each keyword capture the most profit for me? If I set bids to an ROI target of say 50% across my account I will certainly under perform. For some of my low margin adgroups, I will underbid and get no traffic. For my high margin adgroups I will overbid and waste spend. I need a system that tests adgroups and keywords at different levels to find where they perform best.
Profit max = marginal revenue - marginal cost.
Also see the flash video at MediaBoost.com for an excellent description of profit maximizing.
(2) Integration with at least Adwords, YSM, & Adcenter. I advertise across the major platforms. Also, the system must be able to bid discriminate between content and search match.
(3) Weight bidding based on recent data. My revenue for conversions changes cyclically. The bidding system must account for what I tell it a conversion is worth to me now, not itís historic worth.
--- MY RESEARCH ---
Hereís what Iíve found through my research. MediaBoost seems to be the only company that overtly claims a profit maximizing bidding algorithm. Unfortunately, they only work for Adwords. If anyone has a review of MediaBoost Iíd be interested to hear how they perform (Booster, Iíve seen you posting out there and I invite you to chime in!).
Efficient Frontier uses a portfolio-based bidding algorithm. Their system is premised around the idea of maximizing for X (Revenue, Signups, Orders) given a constraint Y(budget). Efrontier then manages your budget across your portfolio of keywords. Iím not interested in maximizing revenue, signups, or orders and I donít have a set budget that I need to manage.
I have also talked to the sales people at DART. Their big claim is that because the search engines share actual bid info with them, they are able to make better decisions based on real data, not estimates. I can see real data creating a competitive advantage, but DART only applied simple rules like ďbid to ROIĒ or ďbid to CPAĒ. Also, I did not care for their interface or syndication process to different ad platforms.
Now Iím starting to research firms like Omniture, ChannelAdvisor, Inceptor, and Bidhero. Anyone tried these agencies? Any other suggestions or links to trusted reviews? I know I might not find my wish list I outlined above, but I canít see shelling out big bucks to have these agencies ďbid to ROI.Ē I can do that.
I know this was a long post.
| 11:20 pm on Dec 5, 2007 (gmt 0)|
I've recently tried and reviewed several bid management tools.
I was less than impressed by MediaBoost even though it was obvious they had smart people behind the algorithms. The execution, while aesthetically pleasing, was largely broken.
SearchForce has been great once you figure out all the intricacies. It's obvious that they need to get a usability expert to consult on their platform. They are thinking about bid management at a very different level than any other company...a level that's more intelligent, but harder to grasp. I definitely recommend evaluating this software.
We've moved on to testing Marin Software. They have a higher price point, but I'm very impressed with how they think about bid management. It's obvious that search marketers built this platform. It does have its issues, but they appear to be minor. I cannot speak to the performance yet, but I've been impressed with what I've seen.
We have several old Efficient Frontier clients and were able to double or triple their performance once we pulled them off that platform. It's a great plug, play, and forget platform if that's what you plan to do. The science is good and works very well for some clients. I'd argue that it's a bad match for others.
Inceptor was pretty interesting. They were a lot like SearchForce, but the customer service wasn't there to keep me interested. That was before they were acquired so maybe it's better now.
Good luck. It's obvious that there's a lot of room for improvement in this industry.
| 3:23 am on Dec 6, 2007 (gmt 0)|
I've been using Efficient Frontier for most of this year after previously using Dart & Omniture, and am impressed with the results so far.
In terms of maximising your profits, EF will maximise your revenue, and you just work out at what point along your spend/revenue graphs profit is maximised - you can do this pretty easily without needing to rely on a bid management tool. In that sense, I can maximise my profits quite easily whilst still using EF & just doing some manual calculations.
| 4:45 am on Dec 6, 2007 (gmt 0)|
This is very interesting and something I've been looking into for some time, for affiliate business though.
Unfortunately, I was not able to find reputable and reliable solution that would be capable of being plugged into major affiliate networks on the other side, in order to connect cost and earnings in one report, on a keyword level.
In regards of such platforms, weíve come across Atlas at some point. Not sure if any of you had any experience with them.
Sorry about not being capable of saying anything more in relation to the initial query.
| 5:29 am on Dec 6, 2007 (gmt 0)|
Great replies so far! Migriffin, I had not come across SearchForce or Marin Software before. After taking a quick look I want to learn more. Iíll call these guys and research it.
I tried Efficient Frontier and they didnít work out for me. The problem with maximizing to revenue is that the system doesnít care all that much if it bids on terms that the system knows will lose money. Because it is a portfolio approach, you will get money losing adgroups lumped in with money making adgroups. Efrontier probably works better for firms who have a goal like, ďWeíve got $10K to spend on PPC this month, letís get as many people to sign for our newsletter as possible.Ē It didnít work for me.
It seems like most of the big name SEMs out there were built with different goals in mind. I wonder if the founders of many of the SEMs ever tried to put together a PPC campaign. It feels like the systems were built to pitch huge companies who think more in terms of ethereal concepts like branding or lifetime customer value.
Keep the comments coming!
| 6:17 am on Dec 6, 2007 (gmt 0)|
It depends on how you use it I guess, and as you said, what your objectives are. We used to use Dart, and became frustrated with the results, whereby the system tries to maximise on a keyword by keyword approach. With EF, whilst some adgroups will lose money, on the whole, they will make more - through the portfolio approach, it makes tradeoffs to get the highest return.
You may have $10k to spend, but you may find that you're actually better off only spending $7k in terms of profit generated. I guess it is a case by case situation though - some companies will want to spend $10k regardless because they have allocated budgets for it, and can't afford not to spend it (they may lose it next year etc)
In terms of profit maximisation, if you work out at which point you break even (100% ROAS etc), you can work backwards in terms of monthly spend to find the point of profit maximisation. Your revenue (for example) curve will be converse shaped, and your cost curve will be concave - the point at which the two are furthest apart should be the point you are spending at to maximise profits - you could apply this methodology to various objectives with a bit of tweaking etc.
Am interested to hear how others have approached this problem, insights etc?
| 1:27 pm on Dec 6, 2007 (gmt 0)|
Contact the makemetop.co.uk folks and ask them for their ppc bid strategies break down, the one you're looking for is in there.
| 3:33 pm on Dec 6, 2007 (gmt 0)|
I think it also depends on your business model. If you are selling only one product, it's very easy. If you are selling many products (or even worse, services) it's getting harder.
The system should at least be able to recognize what you sold and what value has for you that sale. It's clear that some keywords sell mostly cheap products while other - mostly expensive products, but you can't make a rule that the value for one sale is exactly x euro.
| 5:54 pm on Dec 6, 2007 (gmt 0)|
In regards of all solutions that have been mentioned here:
Is it common for all of them to give PPC aggregated reporting? For example, can you create a group of campaigns that are active on Yahoo, Google, and MSN and run a report that will show aggregated results?
| 7:26 pm on Dec 6, 2007 (gmt 0)|
I worked at Efficient Frontier for four years (up until 3 months ago, now I'm onto a new venture) and was one of the first 2 employees there. So I'm still an EF fanboy, but removed enough so that my opinion is somewhat disinterested.
Efficient Frontier can and does do profit optimization for clients, and has been doing so as far back as 2003. It's actually not a common goal among Google & Yahoo's top 1000 customers, but the reasons for that are too long to comment on here.
Over the years I've seen EF's system throw out some pretty huge improvements in campaigns, including profit-optimization campaigns; if you want to sticky me I can walk you through those.
It's not just portfolio algorithms that make EF's system better; it's accurate traffic estimation, something that most other firms don't have, as well as deep knowledge on the right way to structure campaigns to facilitate optimization.
| 4:06 am on Dec 7, 2007 (gmt 0)|
It would be interesting to see how BidHero stacks up - have always been a big fan of ClickTracks but never explored BidHero. SearchForce - we tried to get in touch with someone there for a while but didn't have much luck. Atlas would not be in my consideration set. Maybe MSFT will do something to improve it.
It's been a real challenge to find something that works well.
| 5:12 am on Dec 7, 2007 (gmt 0)|
I just had a webex demo with SearchForce today. I came away with mixed impressions. Like most of the firms, they donít offer or claim a profit maximizing bid algorithm Ė instead they focus on the common metrics, CPA, ROAS, etc. They offer a portfolio bidding strategy, but Iím pretty sure you can opt in or out of the portfolio setting.
I asked the question, ďLetís say I set my account to a 50% ROASÖ I see that as a problem because I canít get 50% out of some adgroups or keywords, so Iíll underbid and get no traffic, while at the same time Iíd do much better in the profit department bidding certain adgroups or keywords to 100% ROI. How does SearchForce deal with this problem?Ē The response was that SearchForce allows you to create sub-portfolios and assign different ROAS values (same for groups of keywords). Problem is that, given a large campaign, it appears difficult to manage this sub-portfolio approach. I would need to manually create a sub-portfolio, then test out different ROI levels to see which yields the most profit, then set it to that number. That is the bulk of what Iím looking to outsource.
Here are the positives I saw for SearchForce. They are flexible when it comes to setting up tracking and reporting. Also, they have simple tools that give you control over bidding aggressiveness and sample size requirements for bid adjustments. Finally, the price seems more reasonable.
Migriffin, you said ďSearchForce has been great once you figure out all the intricacies.Ē Can you shed some light on the intricacies you needed to overcome? Also, what were the bidding goals you were trying to implement?
| 11:13 pm on Dec 7, 2007 (gmt 0)|
Hi enterthevortex. I think SearchForce requires a shift in the way of thinking about bid management. It's quirky and not always intuitive, but it allows a tremendous amount of bulk management and updating. One thing I like is that you can filter by any criteria you want, may a "portfolio" for that, and then bulk change items in that portfolio. So I can manually bulk change keywords across three engines that have spent over $100 in three weeks with a ROAS of 20% or less. It's pretty useful once you think through all the different scenarios. Then, I can apply a bid rule to these keywords. If you're able to filter them by some performance criteria, you should be able to identify what bid rules you want to apply to this keyword set.
So to answer your 50% ROAS versus 100% ROAS problem, you just setup two keyword sets (ie porfolios) and apply different rules to both. One thing that they probably didn't mention is that when you set an ROAS target, that's the minimum target that it's going to try to achieve. After it hits that, it's going to try to maximize your ROAS. Every time you set a goal on their platform, it's going to try to maximize around that goal. You can set an ROI goal and it will try to get you the highest ROI possible subject to that minimum. If you set a traffic goal, it will try to maximize your total traffic subject to your max CPC.
So if you're willing to spend some time learning SearchForce, I think it's a good option. That said, we're evaluating Marin right now and considering switching because SearchForce is a little too much of a "black box" solution for us.
| 4:31 pm on Dec 9, 2007 (gmt 0)|
In today's search engine environment bid management simply isn't enough. Since both Google & Yahoo have an opaque business model, you've got to focus on quality and relevancy management then simple bid management. In order to obtain a higher quality score you need: correlation between keywords (and a keyword generation tool), small and focused adgroups, relevant ads and only then can you begin optimizing using bid management. It is my belief that mediaboost and efficient frontier only focus on the bid management aspect and don't seem to deal with the relevancy.
[edited by: engine at 3:04 pm (utc) on Feb. 19, 2008]
[edit reason] See TOS [webmasterworld.com] [/edit]
| 8:53 pm on Dec 9, 2007 (gmt 0)|
Hey Mr. Rosen,
I agree with much of what you have said. Iím looking for more of a software solution than a full service keyword/adcopy solution.
[edited by: engine at 3:05 pm (utc) on Feb. 19, 2008]
[edit reason] See TOS [webmasterworld.com] [/edit]
| 9:18 pm on Dec 9, 2007 (gmt 0)|
I wasn't sure if I was allowed to name the company (according to the TOS), but the company name is kenshoo and the URL is .............
[edited by: skibum at 3:03 am (utc) on Dec. 13, 2007]
[edit reason] yea, we can't drop our own links :) [/edit]
| 2:22 pm on Dec 10, 2007 (gmt 0)|
So I'm going to come clean right off the bat and say that I work on the Dart Search Team at doubleclick.
First I'd like to say that profit maximization is done through hard work. This means tinkering with your campaign hands on, having an asset or even a team if you have enough traffic to be watching and managing your campaign. I to this day, have not seen a set it and forget it tool out there.
I would admit that our tool does not have the prettiest interface on the market, I however don't think that "easy on the eyes" is a top priority for a Search Marketing Solution.
Our big claim that we use Real Engine data is quite an advantage. We are the only company that has real API integration with all 3 engines. That means you optimization is based upon 100% real data. The minimum descripency disclaimer I've heard from anyone using sample data, graph estimations, or any sort of algorithim is 10%. That means if everything works perfect, your already off by that margin.
Our relationship with the engines goes much deeper. 2 of the 3 Tier 1s use our tool to trafic their own campaigns...the other is attempting the acquire us. We are in fact the only 3rd party tool that the engines have their people train in at all.
Something that hasn't been mentioned is the ability to do spotlight tag optimization. This means the system has the ability to take into account what conversion the words are generating, and is able to optomize of that. Optomization can be at the keyword, the group, or the strategy level, not just the keyword level. This gives you an effective way to both maximize profits, and lets your weight bidding be based on recent data, as you prioritize your conversions.
On the other side, if your not managing a lot of click traffic, if your not hands on / don't have assets to allocate to the tool, or if you are looking for a black box solutios out tool is NOT for you.
Honestly, this forum shouldn't be used as a sales floor, I deleted half the stuff I've typed, I just wanted a clear view of our product to be out there. I'd be happy to discuss the product, the pros and cons, or anything else, feel free to e-mail me at firstname.lastname@example.org.
[edited by: Sylvanrage at 2:23 pm (utc) on Dec. 10, 2007]
| 3:09 pm on Dec 10, 2007 (gmt 0)|
|We are the only company that has real API integration with all 3 engines. |
What exactly do you mean by that?
And in the spirit of openness, let me tell you that I work for Efficient Frontier.
| 3:34 pm on Dec 10, 2007 (gmt 0)|
We are running just a few ads now, we spend around $50 a day. We are a small company and we sell one type of product but there are more than 50,000 variations. In what level ($ spent per day, number of KW etc) it makes sense to start looking into such bid management tools?
| 5:55 pm on Dec 10, 2007 (gmt 0)|
I should have said Full, rather then real.. we constantly work with Google, Yahoo, and MSN to keep full API compliance with their systems.
Full API integration means we are getting real search engine data, and optomizing on it every time it is released by the engines (up to 12 times a day). This means your campaigns are constantly be optomized on real time data. Also this means any reports you run will also reflect real data.
The neatest thing about PPC advertisement is that you are able to track where every penny goes, and know your return on it... why would you want to add data gaps to something like that?
On a seperate note... I think these are some of the questions to ask when evaluating any Search Marketing Solution?
1. How accurate is your date?
2. How will this create workflow efficencies for me (save me time).
3. How does you system handel bid management?
4. Can you system grow with me, how scaleable is it?
5. What sort of support will my in house team get?
[edited by: engine at 3:07 pm (utc) on Feb. 19, 2008]
[edit reason] See TOS [webmasterworld.com] [/edit]
| 5:58 pm on Dec 10, 2007 (gmt 0)|
Great question... at 50 dollars spend a day with a few ads. I don't think it would make sense to use a in house tool. Most of the advantages offered by tools like mine aren't realized till you are looking at 100ks worth of clicks a month, they are mainly bought to improve workflow effeciency.
I would suggest that you either look into a SEM specialized agency, or bring in a dedicated search specialist in your company if your finding that PPC campaigns could use more focus.
| 10:44 pm on Dec 10, 2007 (gmt 0)|
Itís nice to get DARTís perspective in here since itís such a huge management player. First I want to clarify a bit on what Iím looking for. I do not expect to find a tool that allows me to click a button and watch the profit spill in. Iím looking for a tool that uses and optimizes for metrics that are relevant to accomplishing my goals and allows me to easily manage my portfolio. Iím looking to make money off the traffic I purchase. Therefore, optimizing to a metric like revenue or traffic is worthless to me because the correlation to profit is weak. Optimizing to ROI starts to hint at what I want because at least we are talking about making money. But focusing on a certain ROI number will cause us to under perform in terms of real profit. The problem I had with DART, and pretty much all the other SEM tools, is that the bidding rule goals do not match my goals. I think there is a flip side to your statement that ďprofit maximization is done through hard workĒ -- itís that bidding to ROI isnít all that tough. So why should I pay DART a minimum of $100K a year for that?
Second, I was under the impression that DART is privy to bid info only. Unless that bid info is already normalized for quality score on Google and CTR numbers on Yahoo and other platforms, then DART is still very much in the dark.
Finally, I have an industry-wide questionÖ Why doesnít DART, Omniture, or any similar company offer API access to their management platform? Really? Their whole systems are based on being able to integrate efficiently with other platforms through an API, yet they donít extend their clients the power to manage that single platform in an efficient method that can integrate with our internal tools. Managing for profit wouldnít be nearly as much ďhard workĒ if I could change bidding goals through an API instead of a clunky web interface.
| 10:55 pm on Dec 10, 2007 (gmt 0)|
You would set a cpa goals set to certain spotlight tags, and manage them to get the most clicks at the best price. You can target specific conversions.
I don't know your business, and would love to discuss with you what we do, based on what you do. Basicly roi bidding is not that hard on small campaigns...or manageable campaigs. We have a client with over 19million key words. Much less easy..
As to why we don't allow api integration, its expensive to develope, and with our system constantly having to be updated to stay with the tier 1 api's (we are on a one month development cycle) it would mean that the other side would have to be managed that way too...
If I asked really nice would you go ahead and e-mail me and we could discuss this in depth... Dart search is not a solution for everyone...but hey... it might be for you.
| 9:37 am on Dec 12, 2007 (gmt 0)|
Sylvanrage Ė see sticky mail for response.
| 8:05 pm on Dec 12, 2007 (gmt 0)|
Nice thread :-)
| 9:54 pm on Dec 12, 2007 (gmt 0)|
What's the typical spend required for these tools to be cost effective?
Could an advertiser with, say, a few hundred closely related keywords and an annual spend in the $100k range find value?
| 11:16 pm on Dec 12, 2007 (gmt 0)|
Hereís my report on what I saw on the Omniture sales webex. Iíll start by saying that I was impressed (at least that was my first impression). I understand why bulls flash over the screen as Jim Cramer hits the BUY, BUY, BUY button every time Omniture is mentioned. Okay, hereís the meat of what I sawÖ
Omniture has three product brands that fit into SEM management. There PPC management is called SearchCenter, the site analytics is SiteCatalyst, and their multivariate landing page system is the newly acquired Offermatica. All products are priced and sold separately, but they are integrated and can be controlled centrally.
My focus was on SearchCenter, but I learned that in order to integrate the way I wanted to (share revenue and reporting data with bid management), I would need to use SiteCatalyst. Iím okay with that as long as it doesnít move the price point too high.
The positives I saw with SearcCenter include: Integration with all top tier engines, plus many 2nd and 3rd tier engines as well; Self defined bidding rules; Slick intuitive interface; plus too many bells and whistles to list. My focus was on bidding rules, so I tried not to get distracted by all the toys.
Omniture takes the anti-blackbox approach to bidding: they give you the tools to define how to bid instead running it through a mysterious algorithm. Basically, there are a series of ďif this, then than thatĒ boxes that you can fill out to define bidding preferences. They also have all the normal defaults (CPA, ROI, etc). That all sounds great, but I did see one hang-up that might be a deal breaker. I wanted to know about sample size requirements to apply bidding rules. The response was that the system needed a minimum of 3 days worth of data, and a max of 90 days. HmmÖ #*$! does the number of days have anything to do with sample size? I ran into something similar at efrontier where they wanted 21 days during setup before applying their algo (while they were being paid to watch my account of course Ė little tangent). In statistics, you donít measure sample sizes by days, you measure by numbers. I could have one adgroup that gets 1 click in three days and another that gets 1,000 clicks in three days. This is not a small point -- it means that if I canít define sample size requirements, Iíll be left to do manual bid management until Omniture takes over.
So I followed up the webex with a couple questions to find out more. Hereís where I started to cool down on Omniture... Itís been 8 days since I sent over a couple buying questions and they havenít answered any of them (got a quick email that they would need a couple days to answer). Here is the first of five questions I sent over:
--ď(1) 3-day Requirement for bidding rules. We saw that in order to run bidding rules, the system required a minimum of 3 days of historical info. How does this function work? Is this a moving average of the last 3 days, or a starting requirement? What data needs to be collected in the three days? For example, if we uploaded an adgroup but our starting bid was too low and it didn't get any clicks or impressions, would the bidding rules be unlocked on the fourth day? If so, what's the point of the requirement? We are trying to figure out what kind of bid management we'll need to do outside the Omniture platform to get adgroups up and running.Ē
Awaiting a response as my interest cools with each passing dayÖ
| 11:29 pm on Dec 12, 2007 (gmt 0)|
Hey CernyM Ė I think that the fewer number of keywords you are managing the less need there is for a big management company. The reason Iím looking is because Iím trying to manage hundreds of thousands of keywords and thousands of adgroups across multiple platforms. Iím not discouraging you from looking into this solution (and Iím not all that qualified to answer this question), but just keep in mind that with a couple hundred keywords it might just be easy to manage with simple software. SearchForce has the lowest price point Iíve seen so far, so you might give them a look.
| 11:45 pm on Dec 12, 2007 (gmt 0)|
ETV, in Omniture's defense, it makes perfect sense for them to say they'll need 3-90 days of data precisely because of your good point, which is that they can't know in advance of working with you what your data looks like, hence the range. And in EF's defense too (Disclosure: I worked there for 4 years), EF learned over time that it's prudent to set pre-launch data gathering time expectations correctly.
| 11:50 pm on Dec 12, 2007 (gmt 0)|
Buy why # of days? Why not a sample size of X number of visitors and X number of conversions? The # of days seems arbitrary.
| This 47 message thread spans 2 pages: 47 (  2 ) > > |