Friday, July 22, 2011

HubSpot Spreads Its Wings

The folks at HubSpot have been busy this summer, announcing their acquisition of Performable in June and their HubSpot App Marketplace last week. Both events mark a continued expansion of their product.

The company gave me a brief preview of the App Marketplace back in May, when the public beta had just launched. App markets are quite the fashion right now, and HubSpot’s joins the Eloqua AppCloud announced in June. In both cases, what’s really happening is the vendor has published APIs that make it easier for other vendors to build products that integrate with their systems. Such APIs are available to varying degrees for other marketing automation products too, so an app marketplace isn’t quite as huge a leap as it may seem. But marketplaces do make it easier to find compatible applications and, done right, ensure that deployment is very simple.

I couldn’t find a public list of the available HubSpot apps, but their press release cites a connector for Microsoft Dynamics CRM and my notes from May mention custom analytics and shopping cart integration.

Unlike smartphone app stores or AppExchange, the marketing automation vendor app markets won’t establish their products as “platforms” for a broad range of tasks (although the vendors can dream). They instead extend the functionality of the core product and, mostly, simplify integration with other products that already exist independently. In other words, the app markets are useful but not huge strategic differentiators.

Acquisitions, on the other hand, can be strategically decisive. HubSpot didn’t make that claim for its Performable deal, which shows an admirable sense of reality. Performable offers some interesting capabilities but nothing that radically alters HubSpot’s market position.

The main feature of Performable is an ability to define “events”, which can be page visits, form submissions, or other Web behaviors. This is a useful extension of standard Web behavior tracking techniques. Like other Web tracking, it requires users to install a small Javascript tag on their Web pages. Performable also has existing connectors with a variety of social media, help desk, email, billing, chat and CRM systems. Events and connectors make it easy to build a central database of customer behavior.

Performable leverages this central database with some impressive reporting, showing the first, last, and intervening sources (i.e., the Web site they came from) for visitors who reach each event. Although HubSpot already had reasonable Web analytics, Performable's ability to incorporate additional external activity is a substantial improvement.

Events can also trigger multi-step campaigns that send an email or call an external URL. The URL calls can including parameters with customer information or other data, providing lightweight integration with nearly any external system. Performable also has an impressive landing page builder that supports a/b testing and can ensure that visitors assigned to a particular test group are treated consistently in later visits.

However, the multi-step campaign engine is quite basic. It allows wait periods and conditional steps, but does not allow grouping to automatically exclude customers who meet one condition from subsequent steps. That's a pretty basic feature, typically used to send different messages to different segments at the same stage of a campaign. Users who wanted to do this would need to write conditions for each step that exclude conditions for previous steps. This can be a pain-staking and error-prone chore.

Multi-step campaigns are a weakness in the existing HubSpot system, so it's disappointing that Performable doesn't provide much help. Somewhat similarly, Performable relies on third-party email systems, so it doesn't directly improve HubSpot’s existing email engine, which also lags competitors.

But HubSpot made clear that the Performable acquisition was as much about getting first-rate development talent as about the product itself. In fact, the entire Performable staff joined HubSpot after the acquisition and Performable CEO David Cancel is now HubSpot’s Chief Product Officer. So in that sense, at least, the acquisition is indeed strategic.

Tuesday, July 12, 2011

B2B Marketing Automation Industry Size and Segments

As I mentioned yesterday, our new B2B Marketing Automation Vendor Selection Tool (VEST) asks vendors to estimate the number of clients in each of four size categories.

This provides an interesting overview of the industry. The segments are defined based on revenue. Installation counts are:

Looking at the raw percentages doesn’t make much sense since businesses in each group are quite different. There’s a strong case to be made that micro-businesses in particular have such different needs that their vendors are not really part of the same industry as the rest of B2B marketing automation. I’ve described those differences in this post and go into them in our Vendor Selection Workbook (different from the VEST, and free on the Raab Guide site.)

But if you do want to consider all these vendors as one industry, the minimum adjustment to make is to account for differences in price. The table below calculates revenues using reasonable assumptions about revenue per client in each segment:

Combined with the previous chart, this shows the micro-business segment represents 61% of clients but just 17% of industry revenues. At the other extreme, large business represents just 6% of clients but 28% of revenue. The small- and mid-size companies are the heart of the industry , with 55% of the revenue from 33% of the clients.

The $257.5 million revenue estimate is reasonable but it excludes revenues from B2B marketing automation vendors not in the VEST report and the B2B revenues of B2C marketing automation firms. So I’d estimate total industry revenue at $325* million for 2011. This represents a 50% growth over my estimate for 2010. That is consistent with the growth rate I reported yesterday.

The figures also shed light on the ever-popular question of penetration rates. The table below shows company counts by revenue range from business list compiler Manta. But not all of these are B2B marketers. Looking at the industry categories, I'd put the estimated market at half the total.

The 26.7% figure for the large company category is clearly too high, but that's easy to explain: big companies have lots of divisions, so many vendors have sold to a little piece of those firms. There’s certainly still plenty of opportunity left. It’s possible the 3% figure for mid-size firms reflects some of this effect as well.

Figures for the first three categories are more intriguing. They're much lower than the usual estimates that 5% to 10% of companies have marketing automation. Either the surveys behind those estimates are incorrect or my market definition is too broad.

It’s probably a bit of each: surveys tend to reach people who have above-average interest in the topic, and my 50% figure is based on categories that could potentially use marketing automation, not the categories that have deployed it so far. A count of the pioneer companies, basically tech and manufacturing industries, would reduce the estimated market to anything from one quarter to one tenth the numbers shown. This would translate to penetration rates of 10% to 30%, which is more in line with current estimates.

But I’d argue that the market is already growing beyond this core group, so the long-term potential is considerably larger. That’s great news – so long as vendors don’t get stuck in the current niche and so long as competitors from the CRM, email, Web software, Web advertising or other industries don’t swoop in and snatch it all away.

*The original version of this post estimated $300 million. On consideration, I raised the estimate to $325 million because
- my revised estimate for 2010 was $225
- the 52% growth rate in the previous post was in number of clients, but growth is faster in the higher-priced segments, so the revenue growth would be higher
- average prices are probably rising a bit in the mid-sized segment and big segments, so revenue would rise faster than client counts
- the client counts were gathered in May and June, so they are not quite mid-year figures

I would have gone higher, but the large-company figures are probably overstated in my estimates because many of the 1,200 installations are small, departmental systems that wouldn't generate anything near $60,000 per year.

Monday, July 11, 2011

B2B Marketing Automation Growth Slowed In First Half of 2011

You know that red-hot B2B marketing automation industry? Don’t look now, but growth is already slowing.

Our just-released update to the B2B Marketing Automation Vendor Selection Toolkit (VEST) shows that client counts grew just over 50% over the year ending in June, compared with nearly 100% growth for the year ending last December. That’s a marked decline, and the pattern is consistent across individual vendors: although some grew faster than others, each grew slower than during the previous period.*

You might think the slower rate is expected because each period starts from a larger base. But it turns out that even the absolute number of new clients fell: about 6,100 were added during the recent period, compared 7,000 during the earlier year. I’ll say that again: fewer new B2B marketing automation systems were sold during the past year than the year ending six months earlier. Ouch.

Here’s the actual data:

As of:Client
Year-Earlier Client CountChange
in Client Count
June 201117,21511,0986,11755%
December 201114,1777,2126,96597%

These figures come from eight vendors including all the industry heavyweights: Infusionsoft, OfficeAutoPilot, HubSpot, Pardot, Marketo, Eloqua, Manticore Technology, and Genius. The report actually covers 17 vendors, but the others either were not in the January edition or didn’t provide accurate year-earlier information. The eight companies account for more than 90% of the total installations, so the exclusions are statistically insignificant.**

One obvious question is whether different segments of the industry are growing at different rates. The new report sheds light on this as well. We now ask vendors to estimate their client counts based on four segments:

- micro-businesses, under $5 million in revenue;
- small businesses, $5 to $20 million revenue;
- mid-size business, $20 to $500 million revenue, and
- large business, $500 million or more revenue.

The micro-business segment is concentrated among three vendors: Infusionsoft and OfficeAutoPilot, which serve micro-businesses almost exclusively, and HubSpot, which estimates 50% of its clients are micro-businesses. The remaining five vendors in my data (Pardot, Marketo, Eloqua, Manticore Technology, and Genius) have 69% of their clients in the small and mid-size segments.

The slowdown in growth rates applies to the both sets of vendors, although the small and mid-size group is slightly stronger. Client counts show the same pattern: the absolute increase in the most recent period was lower for the micro-business vendors (4,777 vs. 5,650), while it was essentially flat for the small and mid-size business vendors (1,315 vs. 1,340).

Year-on-Year Growth Rate (Client Count)
Year Ending:Infusionsoft, OfficeAutoPilot, HubSpotPardot, Marketo, Eloqua, Manticore Technology, GeniusAll Vendors Combined
June 201152%68%55%
December 201197%93%97%

So, what does this mean? Is the marketing automation bubble about to burst?

Not necessarily. Year-on-year growth of 50% is nothing to sneeze at, and, as I mentioned earlier, some vendors are growing much faster. Also bear in mind that several vendors have recently received large infusions of funding, which they'll spend on sales and marketing to further accelerate growth.

But it’s still worth sounding a note of caution. Business plans predicated on the industry continuing to grow exponentially now look more dubious than ever. B2B marketing automation could still stall – as B2C marketing automation did – as a niche product for an elite group of sophisticated marketers. It's fine for vendors to expand their product scope, as several are. But they shouldn’t let this distract them from the more fundamental task of growing the base market through promotion, education, and training.

I like irony as much as anyone, but if the demand generation industry failed to generate demand for its own product, no one would be laughing.

*As best we can tell. Some vendors provided partial information, so we had to do some interpolation. And the data flowed in over a two month period, so it doesn’t all align precisely with the January and December time-frames. But the pattern is so strong and so consistent that the general conclusions seem reliable.

**There a few mid-sized vendors who didn’t make the report at all, including Act-On Software and ActiveConversion, which have about 300 clients each. I’d guess these and other vendors add 1,000 to 2,000 to the total client count.