Startup Advertising Spend Calculator

November 22, 2010 - Get free updates of new posts here

I was considering advertising our latest AppSumo Bad Ass developer bundle on Stack Overflow. I called them up and was told the CPMS were $6.25 and minimum spend was $1,000.

I realized I need a way to make decisions about whether that’s a good deal or not.

Here’s the spreadsheet I used based on their data. I figured you could use it as well in deciding which ad campaigns are the best for you.

Get it for yourself.

Key things to always ask when doing your advertising.

  • What is the ctr of the ads?
  • What is the unique visitor to impression ratio? 100,000 impressions doesn’t mean 100,000 saw it.

Leave a Reply

Your email address will not be published. Required fields are marked *

25 responses to “Startup Advertising Spend Calculator

  1. Berry Reply

    Super handy.

    I almost spent $1,200 to advertise in an RSS feed. Turns out most feed readers don’t display their ad. So their claim of 9,000 readers may be true, but that’s not how many would even SEE my ad…

  2. Ankur Jain Reply


    I didn’t understand how you calculated the return (-$2.79) in your spreadsheet above.

    Here is what I follow from your numbers:

    Cost of ad campaign: $1000
    Number of converted user on your site: 0.86
    Expected revenue / converted user: $14.63 (from your spreadsheet)
    Net revenue: $12.6 (=0.86*$14.63)

    If I understand correctly, net ROI will be +$12.60 -$1000 = – $987.4. Am I missing something here?


  3. Jos3ph Reply

    Users == Paying Customers, right?

    Maybe I’m missing something in the way you break this down, but if you spend $1000 & Get 58 visitors you’ve spent $17/visitor like the chart shows.

    If 1.5% of visitors convert to sales, that’s 1-2 sales, & B16 is expected revenue rather than expected revenue per user. Where am I wrong?

  4. Noah Reply

    ctr is click through rate so the % of people that click on your ad.

    CPM is cost per thousand. so if your ad gets shown 1000 times (NOT to 1000 people) how much that will cost you.

  5. Tracy Hall Reply

    Ankur and Jo3eph are correct in concept, if not in calculation. For “each” $1000 spent on this campaign, given your numbers, you will realize 0.86 paying customers – or put another way, it would take a $1161.67 ad spend to expect to gain ***one*** paying customer.

    This makes your rate of return $17 – $1161.67 per new customer – or a loss of ($1144.67) PER NEW CUSTOMER.

    Running your numbers backward (i.e. starting from $17 revenue per customer, 1.5% conversion rate), you can only afford at BREAK EVEN to spend $0.26 per click-through, which means a CPM of $0.09 – which means your $1000 could have over 10,000,000 impressions to garner 57 paying customers about $1000 revenue…

    I could send you the spreadsheet snippet, if you like…

  6. Noah Reply


    Nice catch. On a per customer basis it works but on an overall campaign basis we would need to add in details about # to break-even. I sent you access to the google doc. Feel free to update and I’m glad we have some readers from MIT 🙂

  7. Tracy Hall Reply


    I’ve updated the spreadsheet, including modifying your portion to show the “return on campaign”, as well as calculating impressions from spend and CPM.

    I also added the “breakeven CPM” calculation, using the revenue-per-paying-customer, conversion rate, CTR and impressions/user

    As a user/business owner, I would (and have) run the break-even CPM first, then looked for opportunities cheaper. **IF** you cannot find CPM cheaper for your target demographic, then your only recourse is either (a) increase your conversion (noting a limit of 100% on that 😀 ) or (b) your revenue/paying customer. Simple as that.

    Tracy Hall

  8. Justin Thiele Reply

    CTR is calculated on the total number of impressions not the number of people. So it’s not accurate to take:

    (Impressions/Pageviews per persion) * CTR

    to find New Visitors

    Instead you should just do:

    Impressions * CTR = New Visitors

    Assuming there aren’t many people committing click fraud, this should be fairly accurate.

    This changes the acquisition to 7.2 new paying customers instead of 1.03.

    That puts your cost of a new customer at ($1000/7.2) = $138, which is still alot.

    To make it worth it, you’d need to figure out how to make an ad that attracts a higher CTR on StackOverflow and increase the conversion rate on AppSumo.

    Of course, this doesn’t take in account the lifetime value of a customer.

  9. Wickram Reply

    Hi Noah,

    You can try other advertising options. You can attract more visitors to your blog by advertising on Google content network. Select your list of sites using Google ad planner & use demographics targeting to get max roi for minimum spends. Experiment with $200 on Facebook ads. I’m sure you can get maximum roi by using following demographics targeting.

    Age : 25 -35
    Interests : Stackoverflow, mention few other core devloping sites
    Education : Computer Engineering
    Location : San Francisco


  10. Espree Reply

    This is awesome! Thanks. And thanks Tracy for adding to it as well 🙂

    Perfect timing. Last week I was working on this mind map for myself in to decide which project I want to focus on in the new year – Heres the link –

    It’s a basic step by step breakdown of how to know how much a project idea will monetize before going full force. I got it from a report Tim Ferriss created so it is definitely not my own thinking. Just being that numbers isn’t my thing I had to write it out in an action based methodology for me to be able to implement. aka mind map 🙂 Hope it helps.

  11. Brian Reply

    Hi, Brian here again.

    I’ve just read about the ‘marketing spend spreadsheet’ and while I can only think what a generous person you are, with the very best of intentions, I wonder if you will allow me to offer your readers a couple of Excel-based auto-calculators that I developed for my own business. One will show the true cost of any marketing spend you have done in the past (and the results are terrifying when seen this way) while the other one is a Planning Tool so you can calculate to spend a balanced sum and know what to expect back.
    I will send them to Noah and hopefully he will post them up for free distribution.
    I’m planning an auto break-even projector soon too, which should be handy for newbies planning a new business I would think..
    Hope these can get to you and are found useful…I would love to get some feedback.
    cheers, Brian

  12. FrancoB411 Reply

    Hey Noah, useful spreadsheet. To clarify it a little, I reformatted.

    Yellow = inputs
    CPM field is now yellow
    Return on CPC, CPM, and “new New User” are color coded so when the numbers turn green, you should be good.
    Finally, each ad vendor in on one column, to facilitate comparison shopping.
    Threw some dummy data in there as an example.

    If you’d like to incorporate these changes, drop me an email.



  13. Dan Reply

    To add to my previous comment, I guess the key here is whether the CTR that the given advertiser is quoting refers to:
    – % of clicks over ad impressions;
    – % of clicks over visitors.
    As you rightly point out, you can have more impressions per single visitors, and therefore the result can be massively different.

  14. Stephen Kellett Reply

    Thanks for sharing this Noah. We too approached StackOverflow some time ago to advertise our software tools there. They offered us different pricing to you and the result was the same.

    To make any return (and that would assume our operating costs were zero) on the advertising we’d need to sell each of our tools for over $700. But if you look at most software tools they are in the $200 to $500 range.

    I wrote them a letter explaining the economics and passed it to them via Dharmesh (one of their investors). Never did get a reply. They did not get our advertising.

    I thought we’d been given a bad deal but if they are also offering you a deal that doesn’t work that makes me feel a bit better. Not as if you’re not a known person, whereas no one knows me.

  15. Kristian Rickert Reply

    If you know where to target your customers on which websites, consider spending extra money to acquire a customer.

    If you have a website that attributes the ability to have a returning customer as part of your business model (it should), then most places spend more money than they would make on a first visit to acquire the customer.

    There’s two good things that can come out of this:

    1) Customers. Duh.
    2) if you keep track of how much you spent to acquire a customer, you might be eligible to write it off as a asset. Just be realistic about this cost when you do it and you should be OK (I’m not a tax guy though – so verify this with your controller/accountant).

    The *HARDEST* part about making a successful website is to keep your customers coming back and stay on your site. I’ve seen some of the best websites ever run with amazing tech with amazing services to offer at a low price but their marketing dollars were minimal. This causes a high cost business with kick ass tech and no customers.

    It’s a disease of a first startup – spend money to make it. Pay to get real customers. It took amazon well over 5 years before they saw anything!

  16. Jeannie Reply

    Love reading your diary – bummer though, for some reason the link to the spreadsheet isn’t functioning for me. Any chance you could send the link by email?

    Isn’t it great when you discover that the things you got in trouble for as a kid actually gain respect and admiration when you are an adult?

    To thyself be true – you inspire the rest of us

  17. Steve Sharp Reply

    Thank you Jeannie for sending me the tool. Not sure how you found this post from 2010. When I downloaded the tool it was a little different than the one on the blog so not sure if it is an updated version or what. A tool like this or some version of it will come in handy thought so thanks again. I guess I’ll signup for the newsletter and see what other tools they offer.

  18. Tracy Hall Reply

    Reviving ancient history, because I keep getting questions (and apparently this spreadsheet is now being used as an examples in classes)…

    I’ve updated the spreadsheet, mostly in formatting, to try to help people comprehend it on their own… I’ve labelled (formatted, really) all the user inputs, and have added “name cells/ranges” to allow formulas to read a bit more like english [e.g. ((1000*AverageValueNewVisitor)*CTR)/UniquePerPerson ]

    Hope this helps. Feel free to take and modify at will – this was just to get you started on thinking the right questions, not to create answers.