Teleflora – Site Design

So I just sent flowers online via Teleflora, and I just wanted to comment on what a great site design and layout they have. I had seen their logo in several florists before but didn’t know much about them – I’d always just thought for that type of thing, or the local florist for local stuff.

Their site design and marketing/conversion approach its truly great. Here’s why:

  • Clean layout
  • Products stand out, not the design itself
  • Great knowledge of conversion process – they show you only what you need when you need it. They don’t bore you with showing every single navigation option and link when you are stage 3 of the checkout process and clearly don’t want that.
  • Lots of great sales tools – add person to your contacts list, get yourself reminded of the occasion next year, sample notes to put on the cards (great for a guy like me who might not feel know how to express a nice warm message in such few characters), etc.
  • Great navigation – browse by occasion or category (type of arrangement or product), and the occasions themselves have some crossover

Regarding that last bullet above, the wife instructed me to look for “something springy” for a birthday gift. So which do I use… “Spring” or “Birthday” as the occasion? I know (and care) zippy about flowers. Thankfully there were a few spring birthday arrangements. Nice.

The checkout process was easy too. They also did a great job of removing all unnecessary navigation and info to keep me focused on completing that checkout.

I truly love their design too. I think the best design if you are selling anything that has some sort of visual lure (flowers, clothes, cars, some electronics – anything where aesthetics are a factor) is to have a very simple clean white design that really lets the colors and images stand out. They’ve clearly done that. Okay now no more free publicity for them!

Online Lead Generation – High Quality & Low Quality Leads

My firm deals mostly with service firms. We help to leverage their website and (to a lesser degree) other internet channels to help with their overall marketing efforts. In short, we try and get them more leads and more sales via the web.

With an e-commerce site the conversion process is rather simple.

1) Traffic comes to the website.
2) They either buy or don’t buy.

With a service company, the site’s purpose is typically to generate inquiries or leads, since services don’t frequently completely sell themselves on a website with no personal human interaction. Thus, the process is:

1) Traffic comes to the website.
2) They either become a lead or they don’t (by contacting the company).
3) They either buy or they don’t.

That’s one more step, and its an important one. I once heard something saying that each time you make a person click a link you’ll lose 50% of your traffic. I think that was more true in the early days of the internet, but probably still holds at least some weight. In the lead-generation model there’s one more step where you’ll lose someone.

While I love to count the leads my clients receive, I realize the quantity of leads is meaningless to them. The quantity of quality leads, however, matters a great deal. It is the same for my business – if I can get 3 sales off of 4 leads that is much preferred to 3 sales off of 20 leads. It saves me time, and as we know – time is money.

Thus, the quest is getting high quality leads – those that are more likely to convert to actual sales. If you run an e-commerce site this is similar to the quest for traffic. Quantity is fine, but what you really covet is the high quality traffic that is more likely to convert.

Anyhow, what got me onto this is a new side-venture I have going, a website that serves to help homeowners find local home improvement professionals in Maryland and Virginia. The general service model for home improvement lead generation sites is to get a lead and sell it to a half-dozen firms. The problem is that these are then low-quality leads, because now at least 5 of your competitors also have the lead. As a best case, the lead-to-sale conversion rate for the group would be 1 in 6, or just under 17%.

Our solution – aka the business model of this new venture – is to match customers one-to-one with contractors. Only sell the lead to one contractor. “But wait Jon, that means you’ll make less money!” Eh, maybe initially. But I believe the users and the contractors will be happier with the service. I’ve learned that happy customers lead to good things.

AATH Press Release

All Around the Home

HitsLink – Prevent Duplicate Leads & Conversions From Being Reported


I’m a big fan of HitsLink – a third-party hosted client-side web stats service. Its similar to WebTrends, ClickTracks, Urchin, and all the others – although generally I’d classify it as a more basic offering. For many sites that makes it preferred – as its very quick to setup, is fairly flexible, and gives lots of great data. The new WebTrends 7.5 (or whatever version they are up to now) and other well-known web traffic analysis packages often can go a good bit deeper than HitsLink, but in my opinion the increased price and setup time they come with often does not justify it. I generally find that HitsLink has plenty of data for me – I barely have time to act on it… Lord knows what I would do with more data!

(Main Part of Post)

Anyhow, one issue I’ve had on many sites for a while now is that I had been getting duplicate conversions recorded in HitsLink for the same lead. Essentially it was when someone would fill out a form and maybe accidentially or (due to impatience) click “submit” twice. Or perhaps they’d fill out the form and submit, and then realize they wanted to add or correct something, so they’d hit the back button and submit the form again with a slight modification. One additional cause could be if you are using a “confirmation” page to call the conversion script, and the user decides to stick on the site for a while, maybe either refresh the confirmation page or navigate away, then hit the back button a number of times to find a previous page – during which time they would pass the confirmation page again, causing the script to be called and thus record another duplicate.

For most e-commerce sites you can plug in a unique order ID in the HitsLink conversion script:

… where you would just replace “YOUR-UNIQUE-ID” with a dynamic value generated by your e-commerce system – perhaps a sales order number or something like that. Each number is unique to that one transaction, and thus even if the script/image call was activated several times it would only count 1 conversion for each unique ID.

That’s great for e-commerce sites, but most of my clients are service businesses. They are lead generation sites. The conversion is a lead form that is not connected to any e-commerce engine, so I had no unique ID number available. Then I found a rather obvious solution. Use what you do have. Most contact forms ask for a Name, Email Address, perhaps a Phone Number, etc. Use one of those fields and pass that through. Thus, if submitted the form and was therefore one lead, but activated the script 3 times (see above possible reasons for this), than if you used their email address – as the unique ID it would only register the first conversion and not skew your stats. If you think about it, for most sites an email address will be unique for our purposes, unless of course your site aims to get repeat leads from the same person over time – in which case you should completely ignore this article πŸ™‚

I was elated to discover this method. I had wanted to rely on HitsLink to count leads for my sites, but I often found duplicates skewed the numbers so I’d have to manually count leads or rely on another data source.

There’s also another great benefit. Using either the email address or name or phone number makes it easier to match up the leads in your “Latest Transactions” report with what comes through either to your database or email if your contact form sends to one of those. In the past I tried to match up leads and see where each specific one (on a micro level) came from, what they searched on if they came from a search engine, etc. I had to use the time field to do this. I’d see what time the lead came in via email, and then match that up with HitsLink’s recorded time. Sometimes due to time zones or whatever else the two times might be 1-3 hours off (daylight savings, time zones, etc.) and/or also several minutes off. Thus Lead A might show 10:35 am in my email and 11:29 am in HitsLink. That’s a pain. Matching email addresses or actual names is much easier, and allows you to dig deeper to see where the “good” leads came from rather than just relying on drawing assumptions based on the macro-data.

Okay, So How?

Most forms post to some script or other page – maybe a Pearl page, ASP page, PHP page, etc. to process the form. Then, its common practice to redirect the user to a confirmation page. The processing page runs so quickly and is not displayed that to the user it just looks like they click submit and are taken to a confirmation page.

This is the method I used, and thus what I’ll describe. You can do it other ways too. Steps for this method:

1) Pick your unique ID field and make sure its going to be unique for each lead – a Name (usually unique, although you might have issues with real common names and high traffic sites), an email address (generally pretty unique), etc.

2) On your form processing page you’ll need to add a call for this value into the URL of the confirmation page. Thus, instead of redirecting to you’ll need to redirect to where “unique” can be any parameter name you want and “” must be a dynamic value – a call for the name of the field. In ASP you might try plugging in


into your URL. Something like this would work:

response.redirect (“”&request.QueryString(“name_of_form_field_youre_using”))

Be aware that you may need to play with the syntax as your dynamic call might have apostrophes and parenthesis that cause issues. Talk to someone who knows the programming language and they should be able to give you a work around for the syntax. Once implemented, this will redirect the person to the confirmation page upon form submission, and in the URL it will pass your unique value.

3) The next step is to grab that unique value from the URL and plug it into HitsLinks confirmation image call – which is basically what records the event in HitsLink. Thus you’ll modify that script:

by plugging in the value you passed in the URL into the YOUR-UNIQUE-ID area (replace it). In our example that would look like this:


if you were to use ASP – where < %=Request.QueryString("unique")%> is the ASP call to grab the “unique” value from the URL string. That will render first server-side, such that when the HTML is loaded into the browser it will replace < %=Request.QueryString("unique")%> with or whatever you chose, and record that into HitsLink.

Now you’ve just eliminated the chance for having your HitsLink skewed with duplicate leads from, and you’ve also made it easier to match up the lead that was delivered to you with their campaign, search engine, search phrase, or whatever other data you want in the “Latest Transactions” report in HitsLink.

Wow that took a long time to write! I hope this helps someone or several someones!

My Other Posts About HitsLink:

By the way, I just did a Google search on hitslink and saw that their URL displayed is showing a source=Alexa parameter. In practice, that would make it look as though every visitor that actually came to their site from Google with that search (their name) was recorded as being from the “Alexa” campaign.

As a general rule of thumb I advise ONLY using source= or other campaign tracking codes on links that won’t be indexed – such as when imbedded in a script via a banner ad or on a pay-per-click ad. Those URLs aren’t indexed, and thus won’t skew your stats. HitsLink is great at making a good product, though not so sure they are great at using their own product!!! πŸ™‚

PPC Landing Page No-No

I’m on a roll today – lots of posts πŸ™‚

So I was just doing an “actual” search, as I’m in need of some pre-printed 1099 forms. So of course, I did a search. I found Staples’ site and it had a targeted title, so I clicked it. I’m then taken to this page immediately after searching for “1099 preprinted forms” and clicking on their reasonably well targeted ad:

Poor PPC Targeting

Notice there’s absolutely no mention of 1099 preprinted forms or even tax-related forms anywhere on that page. Instead they want my zip code. My thoughts: Show me you have the product I’m looking for before I’m willing to give you any information about me! I can’t help but think they’d improve their conversion rate a great deal by moving that zip code request page further back in the process. Pull me in a bit more guys before asking my lazy and impatient self (i.e. typical web surfer) for some information. Let me know I’m in the right place!

Super Bowl & Internet Traffic

I just noticed an interesting article on MSNBC about Super Bowl TV Ads (selling this year for $2.5 million for a 30 second spot) and how they relate to traffic spikes on the websites of some advertisers. The article mentions that “microsites” are now becoming a key component of these commercials – enabling advertisers to extend the duration of their exposure and dialog with the viewer beyond that 30 seconds, so long as the viewer feels compelled enough to visit the microsite.

Can we say “call to action”? Especially for more expensive goods, ones the may require a bit more research before a consumer looks to purchase… well… for those I think this is quite effective. I’m not likely to run out right after the game and buy a Buick. But I may visit the website if my computer is in the other room (and given that its already booted up) and check out Buick’s website… That’s another “touch”, and likely to be a much more valuable one than the 30 second spot – b/c this time I initiated it, and I am more likely to focus my visit on the two or three things I’m most concerned with.

Hopefully these sites are done well and have some sort of immediately-actionable call to action themselves!

FireFox, IE and Stupid Design & Site Access Decisions

Sunday is football day for me, so posting on Sunday is rare… but this is really a hoot:

I was looking through my referral stats for this site and noticed a few people found my post about the software scam. As I do on occassion, I went to Google and ran a quick search for one of the referring searches, just to see how the site ranked. It was 3rd or 4th for the particular search – doesn’t matter for our purposes here. Just below my site was another site that is pulling a feed from my site, I clicked on their link to see what they had, as I had not heard of that site before. This is what I got:

Now I love FireFox as much as the next guy, but that said I do probably split my internet surfing 50/50 with FireFox and Internet Explorer. For me FireFox takes and extra minute to load on my PC, and sometimes I’m just not that patient. But still – check out general internet usage stats and even though FireFox has really burst onto the scene, you’ll still see that 90% of people are still using MS Internet Explorer. FireFox gets roughly 8-10% with a very small percentage (maybe 2% or so) for Opera, Safari and other browsers.

Why in the world would you design a site that is only compatible with FireFox? You’re missing 90% of the market!!!

Come on people! Shoot I’ll even give you that with that SEO-targeted site maybe that market is not like the general market – and perhaps you’ll even get a 50/50 split between IE and FireFox on that site – but still, why make the site useless for literally half or more of your audience? That’s about as ridiculous a thing I can think of.

And even if you did do that, at least let the site run with whatever small quirks it might have in IE – maybe it will look a little funny or something, but to allow absolutely no way to view the site at all using the browser that the vast majority of people use (whether you like it or not – whether its a good browser or not, you can’t debate that the market is using it) is just absolutely absurd!

My advice: Do not put up roadblocks that prevent 90% of the universe from accessing your site. If you have such a roadblock, try removing it and you’ll likely see your conversion rate skyrocket!

As a footnote, it seems as though the discounter online/ scam post, the video professor scam post and my MySpace selling for $580 Million posts have drawn likely 75% of my traffic the past month or so. Shows that controversy and community are where the online traffic is at. No real surprise there.

Free Shipping, Order Size and Conversion Rates

A particular site I work with sells products that range in price from $10 to $150. The site does a decent bit of volume, and we have a fair amount of historical data, and fairly steady data at that. That is to say, that unless there is a clear change in the site, dramatic change in traffic or external environment, we typically know what to expect in terms of our weekly conversion rate, total traffic, total sales and average order size. Usually our deviation will be no more than 10% from the expected, and usually much more like 5% or so.

Shipping charges for the products can range a good bit, depending upon the speed of delivery the customer chooses and what product they have. For the most part, we’re not talking about exceptionally heavy or bulky items. It has been my contention that shipping does not play a major role in the buy/don’t buy decision process for the more expensive items, as its then a small percentage of the customer’s total cost – but rather that for the lower priced items (those in the $10-30 range or so) it does play a role. Its entirely possible for someone to buy a $10 product that costs $7 to ship… thus shipping represents a larger relative cost.

So what did we do? We decided to test out the concept of offering free shipping, and see how that impacted our conversion rates and bottom line. Long story short, here’s what we found:

  • Conversion rates did increase with the free shipping offer – about 20% higher or so. Thus, if our rate was 2.0% before, it would have been roughly 2.4% afterwards.
  • Average order size decreased, by about 30%. Thus, we had to ship 30% more orders just to meet the same total revenue as before.

Looking at the above it was clear to us – 20% more orders, but we’d need it to be at least 30% more to make up for the lower order average. My conclusion is that this is simply b/c we removed a barrier for the low-priced items. A $100 item may have cost the customer 10% less in total with free shipping, but the $10 item was now 50% less or more. Thus, we saw alot of very low dollar orders, increasing our conversion rates, but lowering our average order.

Talk to most internet retailers and they’ll tell you that what they want is a way to increase the large dollar orders, not the smaller ones – especially if you are offering free shipping. For a $10 item with free shipping, the shipping itself might cost $3-5, often more than the profit margin (gross margin) on the product itself. You run the risk of losing money. Of course you are banking that this offer also generates more mid and large level orders, thus making it worth your while.

In our experiment we may roughly the same total revenue, slightly less, which resulted from lower average orders but more of them. At the end of the day we decided we’d rather switch back as it was more profitable for us…

However, I do plan to strongly consider following up this small experiment. What really got us was the small orders. What if we picked a price point at which our gross margin would surely cover the cost of offering free shipping, and then some? In this example, perhaps $30 or $40 would work – large enough to not give money away on the low-dollar orders, but hopefully significant enough of an enticement to increase conversions on the mid-range and large dollar orders.

Over-Optimization – What’s the Point?

Sometimes you see a website that is ridiculously over-optimized and you just have to ask… What’s the point?

I can’t take credit for finding this link – as I found it on one of the blogs I read and emailed it to myself, but now I can’t seem to find the article and thus give proper credit. My apologies.

Check that out. Wow. Okay just for kicks lets say that page actually gets traffic. How many people with Migranes do you think are actually going to convert on that site? For that matter, what’s the call to action? What IS a conversion on that site?

Okay just found it – its the link in the body area. So they are banking that someone will do the following, thus making them some money:

1) User finds that page.
2) User doesn’t immediately click away b/c it looks so spammy and worthless.
3) User decides to click the link in the body with no compelling reason to do so.
4) User is taken to homepage and decides to either click on AdSense-like ads or Affiliate links.

Hahaha. Good luck!

Traffic + Conversion = Success

An “SEO” focuses about getting more qualified traffic to a website. A “Conversion Specialist” focuses on doing a better job in producing results with the traffic a site already gets. A successful internet marketer realizes the best case scenario is to address both issues, although most (including myself) will generally advise starting with the scales tipped towards improving conversion, and then gradually tip them more towards generating more traffic, while always making strides in both areas.

An excellent point well worded by Howard Kaplan:

It’s pure lunacy to change your site to accomodate the recommendations of a firm whose stated goals are to provide more qualified traffic, when you’ve previously displayed an utter inability to close on the qualified traffic you currently enjoy. (view post)