How can LIKE '%…' seek on an index?

Question

I would expect these two SELECTs to have the same execution plan and performance. Since there is a leading wildcard on the LIKE, I expect an index scan

Answer 1

These tests (database AdventureWorks2008R2) shows what happens:

SET NOCOUNT ON;
SET STATISTICS IO ON;

PRINT 'Test #1';
SELECT  p.BusinessEntityID, p.LastName
FROM    Person.Person p
WHERE   p.LastName LIKE '%be%';

PRINT 'Test #2';
DECLARE @Pattern NVARCHAR(50);
SET @Pattern=N'%be%';
SELECT  p.BusinessEntityID, p.LastName
FROM    Person.Person p
WHERE   p.LastName LIKE @Pattern;

SET STATISTICS IO OFF;
SET NOCOUNT OFF;

Results:

Test #1
Table 'Person'. Scan count 1, logical reads 106
Test #2
Table 'Person'. Scan count 1, logical reads 106

The results from SET STATISTICS IO shows that LIO are the same. But the execution plans are quite different:

In the first test, SQL Server uses an Index Scan explicit but in the second test SQL Server uses an Index Seek which is an Index Seek - range scan. In the last case SQL Server uses a Compute Scalar operator to generate these values

[Expr1005] = Scalar Operator(LikeRangeStart([@Pattern])), 
[Expr1006] = Scalar Operator(LikeRangeEnd([@Pattern])), 
[Expr1007] = Scalar Operator(LikeRangeInfo([@Pattern]))

and, the Index Seek operator use an Seek Predicate (optimized) for a range scan (LastName > LikeRangeStart AND LastName < LikeRangeEnd) plus another unoptimized Predicate (LastName LIKE @pattern).

How can LIKE '%...' seek on an index?

My answer: it isn't a "real" Index Seek. It's a Index Seek - range scan which, in this case, has the same performance like Index Scan.

Please see, also, the difference between Index Seek and Index Scan (similar debate): So…is it a Seek or a Scan?.

Edit 1: The execution plan for OPTION(RECOMPILE) (see Aaron's recommendation please) shows, also, an Index Scan (instead of Index Seek):