If you were expecting some deep insights on how microbiology can be translated into your Twitter strategy I’m afraid you will be disappointed. However, @microbiology can definitely teach you a thing or two!

I was recently sent some profiles to look at best practice strategy. You would think there would be plenty from fancy social media agencies and big brands. But instead I was sent this @microbiology. These guys are ripping it up on Twitter with 22,000+ followers and plenty of interaction.

Some brilliant techniques they use:

Pinned interactive tweet 

If you are a Twitter advertiser you can pin a tweet to the top of your profile free of charge. What better content to pin than an interactive image from ThingLink http://bit.ly/TDD5IP. They have added links through to buy their book, links to learn more about the author and  even a link to read a chapter as a taster. 

Custom profile images & background

You need to stand out from the crowd on Twitter, especially if you are using promoted accounts (PA). Your ratio from clicks to follow from these ads impacts the price and impression share in a big way. Adding background images helps bring the profile to life. They have also added a header image so that they also stand out on mobile (the majority of Twitter trafic). 

Linking up social channels

Followers on Twitter are more than likely to follow you on other media as well. @microbiology has added a link through to their Facebook profile. If you are paying for Twitter then why not get some Facebook Likes too. The more Likes on Facebook, the more effective stories become. 

Quality sharable content

If you are in the field of microbiology then I’m sure you like nothing more than some juicy pics of insects, bacteria and viruses. The guys provide plenty of that as well as video content and articles. As you would expect, it results in lots of RT’s and engagement. 

Check it out. These guys are worth a follow even if it’s just for Twitter campaign ideas. http://bit.ly/TDEC1I 

I mentored an early stage startup today as they setup their first PPC campaigns. They have a validated business etc, so they are looking at using PPC as a marketing channel to kick everything off and start generating customers. 

Here is a super quick guide to getting that first campaign live and working. If you need any more detail on anything then just ask.

Tracking 

STOP now if you don’t have decent tracking in place. You need to know conversion rates for every keyword or you end up wasting a lot of time and money. It’s easy to setup these days, just use Google Analytics to setup some conversion goals and assign a value. 

It’s basic but you need to calculate your Earnings-Per-Click (EPC). If your EPC is greater than your Cost-Per-Click (CPC) for a keyword you are in business! Big corporates make the mistake of assigning budgets to PPC like TV media. If a keyword is profitable, why on earth would you not run it 100% of the time? Just make sure you know which keywords are profitable on a daily basis. 

Keyword research

Don’t scrimp on your keyword research. There are plenty of free keyword tools out there. Use Google Keyword Tool or similar. 

Jump on to excel and start sorting out all your potential keywords into groups. Each group should be similar in content and meaning e.g “buy wine” and “buy wine online”. You will be targeting these groups with similar ads. “online wine deals” for example would fit into a different group. The goal is to have bold text (where the keyword is present in the ad) for all your keywords. 

Account and campaign structure 

Everyone has their own personal preference here but here is my preferred strategy for launching an AdWords account. 

The strategy

I like to start with a small set of campaigns targeting lots of keywords. The goal is to end up with some insanely targeted high quality campaigns with just 1/2 keywords in each. Agencies won’t normally bother with this because it’s time consuming. But for a startup it can be make or break. So initially start with:

Exact match - Desktop - Search only

- Make sure your campaign settings are targeting users on desktops only and targeting only the search network on google. Your keywords should have [    ] round them so that Google only matches queries that match exactly. 

Broad match - Desktop - Search only

- Make sure your settings are targeting users on desktops only and targeting only the search network on google. Your keywords should be broad match (no brackets or quotes)

Mobile campaigns

If you target mobile customers then clone those campaigns and target mobile only. It’s best to separate things out like this for quality score and ease of management. 

Next steps

Over time you should then look at the performance of the campaigns and pull out the great performers (high CTR and conversion on site). Add them to a new campaign of their own with targeted ads, site links, high bids and watch as your CPC drops! More on the bidding strategy later. 

Account history

Every account, campaign, adgroup and ad has a history. the better the history, the better the performance (lower costs, more clicks). You should remove poorly performing keywords, adgroups or even campaigns from your accounts or things go bad pretty quickly. You don’t really want keywords with less than 1% CTR in there (Unless you are targeting the display network). 

If everything goes Pete Tong, with quality scores of 3 everywhere then think about pausing the whole account and starting again. The same goes for campaigns. You often find the second try that might be identical works much better (Good old Google)! Remember, you lose your valuable history if you do this. 

Quality score

I can’t write an Adwords guide without mentioning quality score. Like Page Rank in organic search it’s a dark horse and an overly used term. In short you need to make sure you are awesome with the below points to ensure your QS is high, and therefore your costs are low with high positions.

1. Landing page 

Make sure you follow basic SEO strategy for the page. Having the keyword in the title, H1, body and some alt tags will help but don’t spam! Basically make sure each landing page is about/closely related to that keyword. It converts better anyway. Also, make sure you have a terms of service and privacy policy. It’s a nice way for Google to check if you are a real site and not just an affiliate bridge page. 

DO NOT create a landing page without links to other pages. Google classes these as bridge pages so will nuke your QS. Make sure the page sits in the normal flow of the site. It’s why things like landing page generating sites don’t work well for Adwords. 

2. Ad relevance

As I mentioned before, each of the keywords in your adgroup needs to match the ads in they show for. 

3. CTR

Your CTR is a huge factor. Bid high at the start to gain rank so that you get a decent CTR. Pause keywords that have less than 1% CTR and target them again with better ads. 

There is a bunch of other stuff but make sure you get these right and it will work well. 

You can check the QS calculation by hovering on the quote icon next to each keyword. Bear in mind that this is not always accurate and takes time to represent the real score. Don’t panic when you first put a campaign live. They can stay at 3 until you are approved etc. 

Aim for a 7. A 10 is worth busting out the champagne. At this point you get site links (make sure you add them in ad extensions) and your CTR skyrockets. With sitelinks it’s very hard for your competitors to remove you from the ad space unless they pay through the nose. 

Adgroups and ads

Make sure you have at least 3 ads in every adgroup. Google loves lots of ads and through rotation you find the ones that work faster. Higher the CTR, the better the QS, the lower your price. 

Try to make the ad stand out from the ad space so look at your competitors. Google told me that a triangle of bold text (keywords matching the ad) works well for example - it does! Things like “UK” in the UK and %’s work too.

Make Sure You Capitalise Your Ad Copy. It’s bad english but gets clicks. 

But every ad space is different, just play around and learn what works. 

Bidding strategy - USE AT YOUR OWN RISK! 

Bid price

I’m aggressive in PPC. I like to start high to get an initial killer QS then reduce the bids. If you can’t afford to do this on all keywords at once then don’t go for weak bids, just reduce the number of keywords but maintain the strategy. Once it’s profitable, move on to the next one. 

Budgets

Again, go high. I am (pretty sure) that Google rewards you for having constant coverage on a keyword. Maybe not deliberately but certainly because of history. If your history beats a competitors things are much easier. If you can’t afford it, reduce target lower volume keywords initially rather than reducing budgets. Get things profitable then go big guns once you are confident things are working.

I hope that helps get you started. There is many years of testing and failing in there but make sure you play around with things yourself. 

If you have any more ideas, questions or hate mail fire away… 

Sorry for a lack of posts recently. I’ve been hiding away for the last 2 months learning to code.

I had a realisation after talking to Gi Fernando (who sold Techlightenment to Experian) a few months ago. In order to stay top of my game I need to code. Gi is a social marketing guru. By social marketing I mean generating REAL value from social, not just posting a few things on a Facebook page. He taps into the Facebook graphs and builds apps with huge hockey-stick growth and millions of users. 

Digital marketing

I see digital marketing in 2 parts (there are probably some fancy agency words here for them):

Creative marketing - Coming up with creative campaigns like Old Spice or Dollar Shave Club. 

Technical marketing - Adwords, SEO, Facebook etc. 

I think you need a mix of both, but technical marketing is where the real value is. It doesn’t matter how creative you get or how much money you spend on production, if you can’t get distribution then it does nothing for your business. 

I’m strong at Adwords and decent at SEO but after talking to Gi I realised that I’ve missed out on the Facebook Graph for too long. Tapping into the graph gives you actual demographic data about your customers as well as access to all their friends for targeted viral marketing. 

Why code?

The problem with the Facebook Graph is that you can’t just fire up a web interface like Adwords and start doing it. Facebook advertising is only a tenth of the power of marketing on Facebook. The magic happens when you deeply integrate with your web app. I won’t go into detail on what can be done with Facebook in this post, but here are a few things that I will talk about in future posts: 

• Facebook login - Get demographic data and add faster signup. 
• Invite functionality - Users market for you. Appears on timelines so 1 invite = multiple visits. 
• Timeline posting - If people use your app and don’t mind sharing then let them. 
• Ticker - The ultimate in realtime peer to peer marketing.
• Story ads - Once you have the graph side working you should be able to generate lots of likes. Follow up with story ads to their friends. 
• Canvas apps - Market your site on the Facebook App Centre, conversion to signup is huge. 

All this magic requires developers, which is why there aren’t many people exploiting it. In order to fully understand the potential, you have to do it yourself. 

So i’ve been learning to code, 2 nights a week and half the weekend. It’s not as hard as it first seems. There is mountains of info out there and modern tech means that lots of the complexity is removed. 

Some super quick tips for beginners like me (I wish I had this list when I started): 

• Learn Rails - This is an amazing tutorial with no experience necessary.  http://ruby.railstutorial.org/ - Don’t copy and paste, get your muscle memory going!
• Setting up your computer - The hardest bit! Find a developer friend to help if you can. The instructions are in the Rails tutorial though. DO NOT give up if this gets hard. After the setup it gets fun. 
• User Heroku for deploying your apps for super simple cheap deployment - Again, no experience necessary - http://www.heroku.com/
• Learn CSS and HTML - this is a beautiful book http://www.htmlandcssbook.com/
• Setup instant Facebook canvas apps with Heroku - A marketers dream - https://devcenter.heroku.com/articles/facebook
• Learn basic Ruby - http://ruby.learncodethehardway.org/
• Stuck? - Just search for it on StackOverflow - So far not had a problem that didn’t have a solution on here. 
• Github - you need an account. Loads of sample code on there to help you learn and you should use this in combination with Heroku. Again, easy to learn and setup. 

The reward

Once you can tap into the graph the reward is huge. I got some help building a site for my mini festival Barbacoustic. It has an invite system through facebook. I sent 10 invites out to friends and got 100 visits in the same day! After a few more weeks learning it shouldn’t be too hard to knock these out without any developer help. 

I hope that all helps. Learning to code has been incredibly rewarding. As a technical marketer I feel like I’ve been shackled by my lack of ability to build web apps properly. Give it a go and let me know how you get on!

I love SEO. I go to SEO conferences, I love talking to black hat SEO’s and I test ideas on my sites. I’m now lucky enough to work in a team that builds multiple startups and I’ve found myself questioning whether SEO will ever be a big part of my marketing focus again. 

Value for money

The lure of SEO is intense. Free traffic! As a 14 year-old I managed to generate 2,000+ visits a day from search engines for free. Not a single penny left my Piggy Bank. These days SEO is expensive and is it worth it for a startup? 

To succeed in SEO you need cash. Google loves big brands, large amounts of great content and lots of links. All three cost money. The return is something you are likely to see in 6 to 12 months time if you are lucky. If you have a brand new domain, which most startups do, then longer. 

Panda’s and Page Rank

No more than a few years ago SEO conferences were filled with the best (and worst) link building and page sculpting strategies. These techniques worked and still work. However, a more common theme these days from top SEO’s seems to be, “build great quality content that people want to share”. 

Has Google finally won the SEO war? In order to win on Google you have to create amazing content that people love. Sure, link building is still effective, but build a page that isn’t duplicate, is high quality and people tweet about and you rank well. 

This just sounds like good product design. Create a well thought out product and make it incredibly easy for people to share online. You can then top it up with some incentivised social marketing or social ads to give it a kick. 

PR

PR & link bait are other “SEO techniques” used to gain links. Again, this isn’t SEO it’s just good content and marketing. Teach your PR and marketing staff basic SEO, the value of links and the rest happens naturally. 

Social plumbing

What I love about working online today is that it’s not just a Google world anymore. I’ve talked to countless site owners who give the same focus to social that they do to search. Referrals from social media are competing with referrals from Google in a big way. Even new entrants like Pinterest are causing a stir in traffic source reports.

The infrastructure is there now and pretty much all content worth talking about should be “social by design”. If you are truly on to something special with your startup then adding the right social infrastructure can deliver far more traffic than search, especially while Google is still working out whether you are a trusted site. 

Focusing on social does not excuse poorly optimised pages

OK, so SEO isn’t dead, it’s just changed. While I think that product and social should come first this doesn’t mean you shouldn’t use best practice SEO while you build your site. On-page SEO is hard for old complex sites, but easy for startups. Get the structure and content right early on. Once your social and product efforts pay off, the organic ranks will follow. 

Get on with the stuff that matters

I’m certainly not the first to say that SEO has changed. I used to spend hours checking Google ranks and the number of inbound links. It mattered before but now decent product and shareability matters more. Focusing on your product and wider marketing strategy rather than losing sleep over how many keyword rich anchor text links your competitor has and your organic ranks will follow. 

AdWords is a great tool for delivering fast targeted traffic to test assumptions about a startup. The likelihood is that your new campaign is targeting keywords or markets that are new to you, so you WILL make mistakes. 

A good AdWords campaign takes time to build up and refine so don’t panic, just make sure you iterate quickly or your measure cycle will get out of hand quickly. You don’t want to find yourself 3/4 weeks down the line with 45 clicks little learnt about your business. 

How to iterate fast on new AdWords campaigns

1. Monitor the campaign every 30/60mins

AdWords looks at a Campaign and AdGroup history when calculating your overall quality score for keyword. If you launch a new campaign make sure you quickly identify poor performing keywords and pause them. You don’t want a few keywords to pollute the rest of your campaign. If you don’t have time to do this then try using Google’s automation rules and pause them if they have a quality score less than 0.5%. 

2. Split up your keywords into niche AdGroups.

This is basic PPC but essential for new campaigns. It helps you quickly set bid prices and monitor performance closer. Again, pause adgroups if they are performing badly. You can always come back to them later. 

3. Refine your ads.

Don’t always give up on bad keywords. Check your ad copy and make sure they are relevant and contain the keyword. Target the Adgroups with low click-through ratios’s first. 

4. Get reporting in place before you launch

It’s an obvious point but easily forgotten. Make sure you have reporting in place to measure your success criteria early. You waste time and money if you are not getting the data you need to validate a hypothesis from the start. 

5. Monitor performance on a keyword level

Keyword level performance is essential for any campaign but important to speed up your learning process. It allows you to focus the campaign but also might give you additional learnings about your business. For example, if “holiday reviews” has a better engagement than “holiday deals” maybe think about the focus of the landing page. 

6. Focus on high performing keywords. 

You probably only need a few hundred clicks to validate an assumption so if you see a high performing keyword on your site then focus in on it. Increase the bid price perhaps, refine the ad and add related keywords. If the traffic is high quality you will validate an assumption much faster. 

I’ve attended Lean Startup Machine @Lean this weekend to hone my Lean ninja skills. The weekend is focused around learning as much as possible about a startup idea in a short period of time over the weekend. 

We wanted to drive PPC traffic to a landing page and measure sign-ups to validate an assumption about the business. We looked at a few solutions including Unbounce.com to build the page. 

In our haste to learn fast over the weekend we forgot to consider Adwords quality score for the campaign and consequently got a brutal Google Adwords slap, ending the test! 

Here is how to get a good landing page quality score on Google AdWords:

Don’t forget basic relevance factors

Although you are mocking up a quick landing page to test an assumption you still need to remember the basics for building a landing page on AdWords for quality score. 

1. Keyword in the H1
2. Keyword in the body
3. Link to a Terms of Service page
4. Link to a Privacy Policy page

Where to host your landing page

Although UnBounce.com is a really nice slick landing page creator it’s own url options should be avoided for PPC. I setup 2 campaigns in the end for the same page. The first used UnBouncepage.com as a url. Google promptly disapproved the ads completely. 

The second campaign was identical in page structure but on our own domain. It got through the Google net ok.

Remember to link internally on your landing page

Although our own URL was approved the ads were approved but the quality score was nuked pretty quickly. This is probably due to having a single page site only. Google’s landing page QS bot Adbot hits your landing page and pulls back some relevant factors to determine quality. According to a Google engineer this includes looking at any pages linked from that page. I’m pretty sure Google will penalise a page with no internal links to other semi relevant pages. It’s a good way for Google to pick up affiliate bridge page spammers. 

So, if you do use a service like UnBounce make sure you launch a number of pages to create a micro site to convince Google you add value to their search results. 

In summary, create a quick microsite rather than a landing page, use your own url and remember to optimise the page to the keywords you are targeting for relevance. 

Do you have any other PPC landing page tips for testing assumptions? 

PPC is one of the simplest ways to dump traffic on your site test to validate a business assumption. I’ve focused on Google AdWords in this post because they command the majority of search traffic in the world so why bother learning multiple platforms to test your ideas through paid search traffic? Just learn AdWords and get testing! It’s also the easiest to get a campaign live quickly. 

Keep tuned for more posts on specific PPC techniques for Lean Startups but here are some ideas to wet your appetite. 

Benefits

Instant traffic

A PPC campaign takes about 10 minutes to setup when you know your way around excel and AdWords. When you hit the big red button to get your ads live it normally takes Google 5 minutes before your ad is being served up on one of Google’s many servers. This drastically reduces your measure cycle. 

Reporting

Google AdWords has built in reporting that can help you measure success quickly and easily. You can test the click-through ratio of your ads to see if people like the sound of your product and you can see conversion rates directly in AdWords when you setup conversion tracking. 

Cheap traffic

Everyone talks about spiralling cost per click costs on AdWords but personally I still think this is an affordable tool most startups. Be lean and limit the budget to only the number of users you need to validate your assumption then turn it off (unless you are already making profit of course!). The cost completely depends on the market you are in. However, there are always some long tail beauties out there to get cheap traffic in for your vertical. Remember you are only running a quick test on a small number of users. 

Things to watch out for

Stay in control

Make sure you learn the basics before jumping in with big budgets. I know a few startups who have chucked on a PPC campaign, gone to bed and woke up £1,000’s poorer. AdWords lets you start with small budgets, so play with small numbers while you learn. 

Keyword intent

Make sure you are targeting the right keywords to validate your assumption. Not all traffic is equal. If you are testing downloads for a mobile app about Britney Spears tour dates, don’t just bid on “Britney Spears” and expect to accurately test your assumption. Think about what people might have in mind when they search for each keyword. “Britney Spears” includes a people looking for their next desktop wall paper. Keep it specific for your initial tests, you can scale later. 

Field of Dreams